JP6471923B2 - Signal processing apparatus and method, and program - Google Patents

Description

  The present technology relates to a signal processing apparatus, method, and program, and more particularly, to a signal processing apparatus, method, and program that can synchronize a plurality of contents acquired through different paths.

  In recent years, devices capable of playing various media contents on the premise of network connection, such as multifunctional mobile phones and tablet terminal devices, are increasing. Further, there is a demand for utilization by cooperation of a plurality of devices using a network function including a conventional television receiver.

  For example, in the cooperation of a plurality of devices, a plurality of media contents having a time synchronization relationship such as the following (A1) to (A4) are received by a plurality of devices through broadcasting, the Internet, etc., and these contents are reproduced in synchronization. An application program is assumed.

(A1) Foreign language audio content, commentary audio content, closed captioning and text information for main video / audio content (A2) Multiple video / audio contents of a musical piece played and photographed for each instrument (A3) One scene Video / audio content taken from multiple angles (A4) Main video / audio content and its high-resolution video / audio content

  It is necessary to reproduce such a plurality of contents in a state where synchronization is maintained at the time of reproduction. For example, as a technology to synchronize multiple contents, extract feature values from each content photographed at the same time by multiple image capture devices, and calculate the similarity of those feature quantities to synchronize multiple contents The technique to make is disclosed (for example, refer patent document 1).

JP 2013-174765 A

  By the way, in reality, when a plurality of devices try to receive media content as described above via different paths, due to factors such as transmission delays, transmission and reception processing delays, and differences in operation clocks of receiving devices. It is difficult to reproduce content while maintaining synchronization. Also, with the technique described in Patent Document 1, when contents to be reproduced in synchronization do not have similar characteristics, the contents cannot be synchronized.

  The present technology has been made in view of such a situation, and makes it possible to synchronize a plurality of contents acquired through different routes.

  A signal processing device according to an aspect of the present technology detects, for each band, a band division unit that divides a sound signal included in first content, and periodic information of the acoustic signal that is band-divided by the band division unit. The periodicity detection unit, the periodicity information integration unit that integrates the periodicity information for each band detected by the periodicity detection unit for all bands, and the periodicity information integrated by the periodicity information integration unit A peak detection unit that detects peak positions and generates peak information; a downsampling unit that uses the peak information of a plurality of time intervals generated by the peak detection unit as information of one time interval; and the downsampling An output unit that outputs information down-sampled by the unit as a synchronization feature amount when synchronizing the first content and the second content to be synchronized.

  A signal processing method or program according to an aspect of the present technology includes a band division process for dividing an acoustic signal included in first content, and a periodicity information of the acoustic signal band-divided by the band division process for each band. Periodicity detection processing, periodicity information integration processing for integrating the periodicity information for each band detected by the periodicity detection processing for all bands, and periodicity integrated by the periodicity information integration processing A peak detection process for detecting a peak position of information and generating peak information; a down-sampling process in which the peak information of a plurality of time sections generated by the peak detection process is information of one time section; The information down-sampled by the down-sampling process is output as a synchronization feature amount when synchronizing the first content and the second content to be synchronized. And an output processing.

  In one aspect of the present technology, the acoustic signal included in the first content is band-divided, the periodicity information of the acoustic signal that has been band-divided is detected for each band, and the periodicity information for each detected band Are integrated for all bands, the peak position of the integrated periodicity information is detected, peak information is generated, and the generated peak information of a plurality of time intervals is set as one time interval information, and downsampling is performed. The information thus output is output as a feature value for synchronization when synchronizing the first content and the second content to be synchronized.

  According to one aspect of the present technology, it is possible to synchronize a plurality of contents acquired through different routes.

  Note that the effects described here are not necessarily limited, and may be any of the effects described in the present disclosure.

It is a figure which shows the structural example of a provision apparatus. It is a figure which shows the structural example of the feature-value calculation part for audio | voice synchronization. It is a figure explaining the down sample of the feature-value for audio | voice synchronization. It is a figure which shows the structural example of a content reproduction system. It is a figure which shows the structural example of the feature-value calculation part for audio | voice synchronization. It is a figure which shows the structural example of a synchronous calculation part. It is a figure explaining the synchronous calculation of the feature-value for audio | voice synchronization. It is a figure explaining the synchronous calculation of the feature-value for audio | voice synchronization. It is a figure explaining the synchronous calculation of the feature-value for audio | voice synchronization. It is a figure explaining the block used as the object of calculation of similarity. It is a figure explaining calculation of similarity. It is a flowchart explaining a transmission process. It is a figure explaining multiplexing of the feature-value for a subchannel signal and audio | voice synchronization. It is a flowchart explaining the feature-value calculation process for audio | voice synchronization. It is a flowchart explaining a main content reproduction process. It is a flowchart explaining a sub content reproduction process. It is a flowchart explaining the feature-value calculation process for audio | voice synchronization. It is a flowchart explaining a synchronous correction information generation process. It is a figure showing an example of application of this art. It is a figure showing an example of application of this art. It is a figure showing an example of application of this art. It is a figure which shows the structural example of a provision apparatus. It is a figure which shows the structural example of a content reproduction system. It is a flowchart explaining a transmission process. It is a flowchart explaining a main content reproduction process. It is a flowchart explaining a sub content reproduction process. It is a flowchart explaining a synchronous correction information generation process. It is a figure explaining the block used as the object of calculation of similarity. It is a figure which shows the structural example of a computer.

  Hereinafter, embodiments to which the present technology is applied will be described with reference to the drawings.

<First Embodiment>
<Features of this technology>
First, features of the present technology will be described.

  The present technology particularly has the following features B1 to B6.

(Feature B1)
According to the present technology, a method and an apparatus having the following configuration for transmitting a plurality of media contents having different contents via different transmission paths and performing automatic synchronization using sound when received by a plurality of different devices are realized. be able to.
(1) The media content is a data stream in which video, audio, image, character information, etc. are multiplexed. (2) The multiple media content to be transmitted is the example shown in the above (A1) to (A4). (3) At least one of a plurality of media contents to be transmitted is defined as a main channel signal, a voice synchronization feature is calculated from the audio signal, and transmission defined by the system The main transmission signal is generated from the main channel signal according to the format. (4) The main channel signal is generated according to the transmission format defined by the system so that the time synchronization relationship between the remaining media contents (sub-channel signal) and the main channel signal matches. Performs multiplexing processing of audio synchronization features and subchannel signals, (5) The main receiving device that receives the main transmission signal outputs the audio signal through a speaker or the like during reproduction of the main channel signal. (6) Includes the feature value for audio synchronization of the main channel signal. The sub-receiving device that receives the sub-sending signal picks up the sound of the main channel signal output from the speaker by the main receiving device using a microphone, etc., calculates the feature value for sound synchronization, and for the sound synchronization of the received main channel signal Automatic synchronization calculation with the feature amount is performed, and synchronization correction information (time difference information) based on the speech feature amount is calculated. (7) Based on the synchronization correction information based on the speech feature amount, the sub-receiving device converts the received subchannel signal into the received subchannel signal. Play with the correction of the synchronization with the main channel signal.

  As the transmission of the data stream (1) above, it is assumed that the media content is transmitted in a network such as a broadcast wave or the Internet, and a logical transmission path occupied by the multiplexed data stream is referred to as a transmission path.

  Further, the above-described “calculation of voice synchronization feature value” and “automatic synchronization calculation” are realized by the technique described in, for example, Japanese Patent Application Laid-Open No. 2013-174765. It is also possible to downsample the audio synchronization feature amount before transmission, or to perform frame rate conversion of the audio synchronization feature amount as necessary during automatic synchronization calculation using the audio synchronization feature amount. .

  By using such a technique, when the sub-receiving device collects the sound of the main channel signal, the automatic synchronization calculation can be performed robustly even in a bad environment where there is noise or noise. Note that this technique is not necessarily used.

(Feature B2)
In the above (feature B1), the transmission side system unilaterally transmits the main transmission signal and the sub transmission signal to the main reception device and the sub reception device, respectively.

  In this case, the sub transmission signal needs to be transmitted prior to the main transmission signal.

(Feature B3)
In the above (feature B1), the sending system unilaterally sends the main sending signal to the main receiving device, and the sub receiving device acquires the sub sending signal via the network or the like at its own timing and performs automatic synchronization calculation. To perform synchronized playback of the subchannel signal.

  As an advantage of this configuration, the sub-receiving device can control the acquisition of the sub-transmission signal for its own convenience, taking into account the transmission delay of the network.

(Feature B4)
In the above (Characteristic B1), the main receiving device acquires the main transmission signal via the network at its own timing and reproduces the main channel signal, and the sub receiving device also receives the sub signal via the network at its own timing. Obtains the transmission signal, performs automatic synchronization calculation, and performs synchronous reproduction of the subchannel signal.

  As an advantage of this configuration, the sub-receiving device can control the acquisition of the sub-transmission signal for its own convenience, taking into account the transmission delay of the network.

(Feature B5)
In the above (feature B1), there are a plurality of audio signals of the main channel signal.

  For example, the main channel signals of a plurality of systems are the main audio and sub audio of bilingual broadcasting. Voice synchronization feature values are calculated for the audio signals of all systems, multiplexed with subchannel signals, and transmitted. The sub-receiving device determines which sound of the main channel signal is being reproduced when performing the synchronization calculation of the collected sound and all the received sound synchronization feature values. The switching of the audio signal output from the main receiving device is also detected by the synchronization calculation.

(Feature B6)
In the above (feature B1), “synchronization deviation” is detected in the automatic synchronization calculation in the sub-receiving device, and real-time correction processing is performed on the sub-receiving device side.

  Since the main receiving device and the sub receiving device operate independently, the audio clock differs and a synchronization shift occurs. Therefore, by detecting and correcting the synchronization deviation, it is possible to reproduce a plurality of contents while maintaining the synchronization.

<Example configuration of the providing device>
Next, specific embodiments to which the present technology is applied will be described.

  First, a configuration example of a providing apparatus that provides content having a time synchronization relationship as in the examples shown in the above-described (A1) to (A4) will be described.

  FIG. 1 is a diagram illustrating a configuration example of a providing apparatus. The providing device 11 includes a main channel signal, which is a signal for reproducing main content (hereinafter referred to as main content), and content (hereinafter referred to as sub-content) having a content relationship with the main content. A subchannel signal which is a signal for reproduction is supplied.

  Here, the main content and the sub-content are at least one of video and audio, and are contents having a time synchronization relationship with each other. That is, at the time of reproduction, it is desirable to reproduce the main content and the sub content in a synchronized state.

  In the following, description will be continued assuming that the main content and the sub-content are each composed of an image signal for reproducing a video and an audio signal accompanying the image signal. Therefore, in this example, the main channel signal and the subchannel signal are composed of an image signal and an audio signal, respectively.

  The providing apparatus 11 includes a conversion unit 21, an output unit 22, a voice synchronization feature amount calculation unit 23, a multiplexing processing unit 24, and an output unit 25.

  The converter 21 converts the supplied main channel signal into a format defined by a predetermined broadcast standard and supplies the main transmission signal obtained as a result to the output unit 22. The output unit 22 transmits the main transmission signal supplied from the conversion unit 21 by broadcast waves, for example, or transmits it via a communication network such as the Internet.

  The voice synchronization feature quantity calculation unit 23 extracts the voice synchronization feature quantity from the voice signal constituting the supplied main channel signal, and supplies the voice synchronization feature quantity to the multiplexing processing unit 24. Here, the audio synchronization feature amount is a feature amount used for reproducing the sub-content in synchronization with the main content when reproducing the main content and the sub-content.

  The multiplexing processing unit 24 adjusts the time synchronization relationship between the audio synchronization feature value from the audio synchronization feature value calculation unit 23 and the supplied subchannel signal, using the supplied main channel signal. That is, in the providing apparatus 11, since the main channel signal and the subchannel signal are in a synchronized state in advance, the multiplexing processing unit 24 uses the main channel signal to perform the audio synchronization feature in a time synchronization relationship. With the amount and the subchannel signal synchronized, the audio synchronization feature amount is associated with the subchannel signal. For example, in the MPEG-4 System, an audio signal, a video signal, etc. are each handled as one media object (ES (Elementary Stream)) and multiplexed. Since the time attribute is defined for each minimum unit called Access Unit (AU) obtained by dividing ES, the audio synchronization feature amount is treated as one media object having time attribute information. Multiplexing with certain media objects is easy.

  Further, the multiplexing processing unit 24 multiplexes the audio synchronization feature quantity and the subchannel signal in a time-synchronized state, then performs format conversion as necessary, and obtains the sub-result obtained as a result. The transmission signal is supplied to the output unit 25.

  The output unit 25 transmits the sub transmission signal supplied from the multiplexing processing unit 24 by, for example, a broadcast wave or transmits it via a communication network such as the Internet. Here, the main transmission signal and the sub transmission signal are transmitted to the content reproduction side system via different transmission paths.

  In the example of FIG. 1, the providing device 11 is configured by a single device, but the providing device 11 may be configured by a plurality of devices, and each process is executed by cloud computing. May be.

<Configuration example of feature calculation unit for voice synchronization>
Further, the voice synchronization feature quantity calculation unit 23 shown in FIG. 1 is configured in more detail as shown in FIG. 2, for example.

  The voice synchronization feature quantity calculation unit 23 includes a frequency band division unit 51, a periodicity detection unit 52-1 through a periodicity detection unit 52-4, a periodicity intensity detection unit 53-1 through a periodicity intensity detection unit 53-4, A periodic information integration unit 54, a peak detection unit 55, and a downsampling unit 56 are included.

  The frequency band dividing unit 51 divides the audio signal constituting the supplied main channel signal into time intervals of about several tens of milliseconds to 100 msec using a window function.

  Here, the processing performed between the frequency band dividing unit 51 and the peak detecting unit 55 is performed for one time interval. At this time, a plurality of time intervals (time frames) continuous in the time direction can be obtained by shifting the time position to which the window function is applied after about several milliseconds to 100 milliseconds. On the other hand, in the downsampling unit 56, the results of a plurality of continuous time intervals are integrated into one, and the feature amount for the new time interval after integration is calculated.

  The frequency band dividing unit 51 divides the audio signal for each time interval into four frequency bands using a plurality of bandpass filters, and the audio signal of each frequency band is converted to the periodicity detection unit 52-1 to the periodicity detection. It supplies to each of the parts 52-4.

  As the bandpass filter, it is effective to use a filter whose pass frequency bandwidth becomes wider as the frequency becomes higher, such as an octave band filter.

  The periodicity detection unit 52-1 to the periodicity detection unit 52-4 calculate the autocorrelation function of the audio signal for each time interval of the predetermined frequency band supplied from the frequency band dividing unit 51, thereby The periodicity information representing the periodicity of is extracted.

  Here, as the periodicity information, the autocorrelation function x (b, τ) itself of the time-delayed audio signal in the frequency band with the index b and the index τ is used, but the autocorrelation function x (b , Τ) divided by x (b, 0) can also be used. As a method for calculating the autocorrelation function x (b, τ), a method using a spectrum peak obtained by performing discrete Fourier transform on an audio signal in a predetermined frequency band can be used.

  The periodicity detection unit 52-1 to the periodicity detection unit 52-4 integrate the periodicity information for each extracted time interval with the periodicity intensity detection unit 53-1 to the periodicity intensity detection unit 53-4. Supplied to the unit 54. Hereinafter, the periodicity detection unit 52-1 to the periodicity detection unit 52-4 are simply referred to as the periodicity detection unit 52 when it is not necessary to distinguish between them.

  The periodicity intensity detecting unit 53-1 to the periodicity intensity detecting unit 53-4 are based on the periodicity information for each time interval supplied from the periodicity detecting unit 52-1 to the periodicity detecting unit 52-4. Calculate the intensity of periodicity for each interval. Specifically, the maximum value of the autocorrelation function x (b, τ), which is periodicity information in τ other than τ = 0, is calculated as the intensity of periodicity. The greater the periodicity strength, the stronger the periodicity of the processing target speech signal, and the smaller the periodicity strength, the more likely the periodicity of the processing target speech signal will be noise.

  The periodic strength detecting unit 53-1 to the periodic strength detecting unit 53-4 binarize the periodic strength for each time interval depending on whether or not the threshold value is exceeded, and use it as periodic strength information for each time interval. . That is, for each time interval, when the intensity of periodicity exceeds a predetermined threshold value, the periodicity intensity information is set to 1, and when the intensity of periodicity is less than or equal to the predetermined threshold value, the periodicity intensity information is set to 0. The The periodic strength detection unit 53-1 to the periodic strength detection unit 53-4 supply the periodic strength information for each time interval to the periodicity information integration unit 54.

  In the following description, the periodic intensity detector 53-1 to the periodic intensity detector 53-4 are also simply referred to as the periodic intensity detector 53 when it is not necessary to distinguish between them.

  Based on the periodicity information for each time interval supplied from the periodicity detection unit 52 and the periodicity information for each time interval supplied from the periodicity intensity detection unit 53, the periodicity information integration unit 54 Perform periodicity integration processing to integrate periodicity information for each section. Specifically, the periodicity information integration unit 54 obtains the sum of the autocorrelation functions x (b, τ), which is periodicity information, for each time interval using the following equation (1).

In Equation (1), N _b represents the total number of frequency bands, and p (b) represents periodic strength information. N _p represents the number of frequency bands in which the periodic intensity information p (b) is 1.

  The periodicity information integration unit 54 supplies the peak detection unit 55 with the sum S (τ) of periodicity information for each time interval obtained as a result of the periodicity integration processing.

Peak detector 55, for each time interval, performs peak detection on the sum of the periodicity information supplied from the periodicity information integrating section 54 S (tau), next to 1 the value of the peak position tau _p, the peak position value other than tau _p generates peak information P (tau) which is a 0. As a peak detection method, for example, there is a method of detecting the index τ when the differential value of the sum S (τ) of periodicity information changes from positive to negative as the peak position τ _p .

The peak detector 55 is the sum of the periodicity information of the peak position τ _p S (τ _p) is smaller than a predetermined threshold, to the peak information P of the peak position τ _p (τ _p) and 0 It may be. Thereby, the noise of the peak information P (τ _p ) can be reduced. The peak information may be the sum S (τ) of periodicity information itself.

  The peak detection unit 55 supplies the peak information P (τ) for each time interval to the down-sampling unit 56 as time series data of the feature amount for voice synchronization for each time interval.

The down-sampling unit 56 integrates the voice synchronization feature quantities of the plurality of time sections supplied from the peak detection unit 55, that is, the peak information P (τ) of the plurality of time sections, as new one time section information, Peak information P ′ _i (τ) is generated as a final audio synchronization feature amount. In other words, the down-sampling unit 56 generates peak information P ′ _i (τ) by down-sampling the peak information P (τ).

In P ′ _i (τ), τ is an index indicating a time delay, and i is an index indicating a time interval. The down-sampling unit 56 supplies the peak information P ′ _i (τ) for each time interval obtained in this way to the multiplexing processing unit 24 as time-sequential data of voice synchronization feature values for each time interval.

Here, generation of peak information P ′ _i (τ) will be described with reference to FIG. In FIG. 3, the vertical axis indicates an index τ indicating time delay, and the horizontal axis indicates time, that is, an index i indicating a time interval.

In this example, a series of peak information P (τ) is shown on the upper side in the figure, and a series of peak information P ′ _i (τ) is shown on the lower side in the figure. In particular, in FIG. 3, the time delay is τ, and the peak information P (τ) of the time interval specified by the index i is represented by P _i (τ). Each square represents the peak information of one time section. In particular, a white square represents that the peak information represented by the square is 0, and a black square represents that the peak information represented by the square is 1.

In the drawing, the length of the time section of the peak information P _i (τ) shown on the upper side is 8 msec. That is, the peak information P _i (τ) is calculated at a time interval of 8 msec. Here, the time delay τ is the same, and the four peak information P _i (τ) adjacent in the time direction (time interval direction) are integrated into one peak information P ′ _i (τ). The Therefore, the time interval of one peak information P ′ _i (τ) is 32 msec.

For example, the down-sampling unit 56 integrates (down-samples) the peak information P _i (τ) by calculating the following equation (2) to obtain peak information P ′ _i (τ).

In the calculation of the formula (2), if there is at least one peak information P _i (τ) whose value is “1” among the four continuous peak information P _i (τ) to be integrated, The value of the obtained peak information P ′ _i (τ) is “1”. Conversely, when the values of the four continuous peak information P _i (τ) to be integrated are all “0”, the value of the peak information P ′ _i (τ) obtained by the integration is set to “0”. The

In this way, by down-sampling by obtaining the logical sum of the peak information P _i (τ) continuously arranged in the time interval direction, the information on the peak position included in the series of peak information arranged in the time direction is down. It will not be removed by the sample. As a result, it is possible to hold how the information regarding the peak position transitions in the time delay direction even after down-sampling.

For example, when the downsampling peak information P _{i (τ),} simply the value of one peak information P _{i (τ)} of a plurality of time intervals of the peak information P _{i (τ),} after down-sampling When it is adopted as the value of the peak information P ′ _i (τ), the amount of information is lost and the accuracy of the synchronization calculation is reduced. That is, the performance deteriorates.

Specifically, for example, four if the peak position is transitioning over time intervals, simply when the peak information P _{'i (τ)} by thinning out the peak information P _{i (tau),} the transition during the one time interval Only the peak information P _i (τ) is adopted as the final feature amount, and the information on the transition of the peak position is lost.

On the other hand, in the method of outputting an appropriate value as the peak information P ′ _i (τ) based on the peak information P _i (τ) of a plurality of time intervals at the time of down-sampling as described above, the time interval after down-sampling Information indicating that a transition has occurred can be held within one hour interval after down-sampling. As a result, detection performance can be maintained even when down-sampling is performed.

  In addition, by performing such down-sampling processing, it is possible to reduce the transmission amount when transmitting the audio synchronization feature amount. In addition, the required capacity can be reduced when the calculated voice synchronization feature amount is held in the memory or storage.

Furthermore, it is possible to reduce the amount of calculation when performing synchronization processing between two audio synchronization feature quantities. In the synchronization processing, when the length of the input feature amount is increased by n times, the calculation amount is increased by n ² times, and thus the effect of the down-sampling processing is great. On the other hand, simply performing the thinning-out process degrades the synchronization detection performance. Therefore, it is necessary to perform a down-sampling process while holding necessary information as in the down-sampling method by the down-sampling unit 56. It becomes.

  In FIG. 3, the example of downsampling the peak information as the feature amount for voice synchronization to 1/4 has been described. However, conversion at any other rate (downsampling) such as 1/2 or 1/8 is performed. It is also possible to do.

  Further, when down-sampling peak information, a method other than the calculation method of Equation (2) described above can be used.

For example, when the value of the peak information P _i (τ) is “1” for two or more time intervals among four time intervals, the value of the peak information P ′ _i (τ) after down-sampling is “1”. It is good. Further, when the value of the peak information P _i (τ) is “1” for three or more time intervals, the value of the peak information P ′ _i (τ) after down-sampling may be set to “1”. When the value of the peak information P _i (τ) in all the time sections is “1”, the value of the peak information P ′ _i (τ) after down-sampling may be set to “1”.

Furthermore, when the value of the peak information P _i (τ) is “1” continuously for two time intervals or more in the four time intervals before down-sampling, the peak information P ′ _i (τ) after down-sampling The value may be “1”, or the value of the peak information P ′ _i (τ) after down-sampling when the value of the peak information P _i (τ) is “1” continuously for three or more time intervals. May be set to “1”.

In the above description, the method of down-sampling the peak information P _i (τ) in the time axis direction (time interval direction) has been described. However, the peak information P _i (τ) is down-sampled in the time delay τ direction. Also good.

In such a case, the down-sampling unit 56 down-samples the peak information P _i (τ) by calculating, for example, the following equation (3) to obtain peak information P ′ _i (τ).

In the calculation of Expression (3), four peak information P _i (τ) arranged in the same time interval continuously arranged in the time delay τ direction is integrated into one peak information P ′ _i (τ).

In this case, among the four successive peak information P _{i (tau)} which is integrated peak information even one value "1" P _{i (tau)} when there is a peak information P obtained by the integration ' The value of _i (τ) is “1”. Conversely, when the values of the four continuous peak information P _i (τ) to be integrated are all “0”, the value of the peak information P ′ _i (τ) obtained by the integration is set to “0”. The

Furthermore, the peak information P _i (τ) may be down-sampled in both the time interval i direction and the time delay τ direction.

In such a case, the down-sampling unit 56 down-samples the peak information P _i (τ) by calculating, for example, the following equation (4) to obtain peak information P ′ _i (τ).

In the calculation of Expression (4), two peak information P _i (τ) of the same time delay τ that are continuously arranged in the direction of the time interval i and the time delay with respect to the two peak information P _i (τ) A total of four pieces of peak information P _i (τ) composed of two pieces of peak information P _i (τ) arranged adjacent to each other in the τ direction are integrated into one peak information P ′ _i (τ).

In this case, among the four peak information P _{i (tau)} to be integrated, if even one value is the peak information P _{i (tau)} is "1", the peak information P obtained by the integration _'i ( The value of τ) is “1”. Conversely, when the values of the four pieces of peak information P _i (τ) to be integrated are all “0”, the value of the peak information P ′ _i (τ) obtained by the integration is “0”.

As described above, when the down-sampling unit 56 down-samples the peak information P (τ) to obtain the peak information P ′ _i (τ), the obtained peak information P ′ _i (τ for each new time section is obtained. ) Is supplied to the multiplexing processing unit 24 as time-series data of voice synchronization feature values for each time interval.

<Example configuration of content playback system>
Next, the configuration of a content reproduction system that receives the main transmission signal and the sub transmission signal transmitted from the providing apparatus 11 as the main reception signal and the sub reception signal, respectively, and reproduces the main content and the sub content will be described. Such a content reproduction system is configured, for example, as shown in FIG.

  The content reproduction system shown in FIG. 4 includes a main receiving device 81, a display unit 82, a speaker 83, a microphone 84, a sub receiving device 85, a display unit 86, and a speaker 87. Here, the case where the content reproduction system is configured by a plurality of devices is shown as an example, but the content reproduction system may be configured by one device.

  The main reception device 81 receives the main reception signal transmitted from the providing device 11 and controls the reproduction of the main content obtained from the main reception signal.

  The main receiving device 81 includes an input unit 111 and a reproduction processing unit 112.

  The input unit 111 receives the main transmission signal transmitted from the providing device 11 as a main reception signal and supplies the main transmission signal to the reproduction processing unit 112. The reproduction processing unit 112 extracts an image signal and an audio signal of the main content included in the main reception signal supplied from the input unit 111, supplies the image signal to the display unit 82 for reproduction, and transmits the audio signal to the speaker. 83 to be played back. That is, the playback processing unit 112 controls the playback of the main content.

  The display unit 82 includes, for example, a liquid crystal display device, and displays the main content image (video) based on the image signal supplied from the reproduction processing unit 112. The speaker 83 is an audio reproduction device, and outputs the main content audio based on the audio signal supplied from the reproduction processing unit 112.

  The microphone 84 picks up the sound of the main content output from the speaker 83, and supplies the sound signal obtained as a result to the sub reception device 85.

  The sub reception device 85 receives the sub transmission signal transmitted from the providing apparatus 11 as a sub reception signal, and controls the reproduction of the sub content obtained from the sub reception signal.

  The sub reception device 85 includes an audio synchronization feature quantity calculation unit 121, a buffer 122, an input unit 123, a separation processing unit 124, a buffer 125, a synchronization calculation unit 126, and a reproduction processing unit 127.

  The voice synchronization feature quantity calculation unit 121 calculates a voice synchronization feature quantity from the voice signal supplied from the microphone 84 and supplies it to the buffer 122. The buffer 122 temporarily records the feature value for voice synchronization supplied from the feature value calculation unit 121 for voice synchronization.

  The input unit 123 receives the sub reception signal transmitted from the providing apparatus 11 and supplies it to the separation processing unit 124. The separation processing unit 124 separates the sub reception signal supplied from the input unit 123 into the audio synchronization feature amount and the subchannel signal, and supplies the separated feature to the buffer 125. The buffer 125 temporarily records the audio synchronization feature amount and the subchannel signal supplied from the separation processing unit 124.

  The synchronization calculation unit 126 synchronizes the main content and the sub-content based on the audio synchronization feature quantity recorded in the buffer 122 and the audio synchronization feature quantity recorded in the buffer 125. Is generated and supplied to the reproduction processing unit 127. That is, the synchronization calculation unit 126 performs matching processing between the audio synchronization feature amount extracted from the audio signal obtained by collecting the sound and the audio synchronization feature amount included in the sub reception signal, and thereby performs the main content and the sub content. Is detected, and synchronization correction information based on the audio feature amount indicating the deviation is generated.

  The reproduction processing unit 127 corrects the reproduction timing (time) of the subchannel signal recorded in the buffer 125 based on the synchronization correction information supplied from the synchronization calculation unit 126, and the image signal and audio as the subchannel signal are corrected. The signals are supplied to the display unit 86 and the speaker 87, respectively. That is, the reproduction processing unit 127 controls the reproduction of the sub content. For example, when the MPEG-4 System is used and the feature value for audio synchronization is handled as one media object, and synchronized and multiplexed with the media object of the subchannel signal, the Access Unit (AU) that is the minimum unit of each media object Since the time attribute is defined for each of (), an appropriate reproduction timing (time) of the media object of the subchannel signal can be calculated from the above-described synchronization correction information.

  The display unit 86 includes, for example, a liquid crystal display device, and displays an image (video) of sub-content based on the image signal supplied from the reproduction processing unit 127. The speaker 87 is an audio reproduction device, and outputs sub-content audio based on the audio signal supplied from the reproduction processing unit 127.

<Configuration example of feature calculation unit for voice synchronization>
Further, the voice synchronization feature quantity calculation unit 121 shown in FIG. 4 is configured in more detail as shown in FIG. 5, for example.

  The voice synchronization feature quantity calculation unit 121 includes a frequency band division unit 151, a periodicity detection unit 152-1 through a periodicity detection unit 152-4, a periodicity intensity detection unit 153-1 through a periodicity intensity detection unit 153-4, A periodic information integration unit 154 and a peak detection unit 155 are included.

  The frequency band dividing unit 151 to the peak detecting unit 155 are the same as the frequency band dividing unit 51 to the peak detecting unit 55 shown in FIG. However, the frequency band dividing unit 151 and the frequency band dividing unit 51 can set the shift time of the window function to different values. For example, when the sub-reception device 85 has abundant computing resources, the frequency band dividing unit 151 uses the shorter shift time, so that it is possible to extract the feature amount for voice synchronization with a finer granularity.

  Hereinafter, when it is not necessary to particularly distinguish the periodicity detection unit 152-1 to the periodicity detection unit 152-4, they are also simply referred to as the periodicity detection unit 152, and the periodicity detection unit 153-1 to the periodicity intensity detection. When it is not necessary to particularly distinguish the unit 153-4, the unit 153-4 is also referred to as a periodic intensity detection unit 153.

<Configuration example of the synchronous calculation unit>
Furthermore, the synchronization calculator 126 shown in FIG. 4 is configured as shown in FIG. 6 in more detail.

  The synchronization calculation unit 126 of FIG. 6 includes a frame rate conversion unit 181, a frame rate conversion unit 182, a block integration unit 183, a block integration unit 184, a similarity calculation unit 185, and an optimal path search unit 186.

  The frame rate conversion unit 181 reads the time series data of the audio synchronization feature amount for each time interval of the main content from the buffer 122, converts the frame rate of the audio synchronization feature amount, and supplies the converted frame rate to the block integration unit 183. Here, the frame rate refers to the number of time intervals per unit time in the time-series data of the audio synchronization feature amount, that is, the length of the time interval.

  The frame rate conversion unit 182 reads the time series data of the audio synchronization feature value for each time interval of the main content from the buffer 125, converts the frame rate of the audio synchronization feature value, and supplies the converted frame rate to the block integration unit 184.

  The audio synchronization feature values held in the buffer 122 and the buffer 125 may differ in frame rate, that is, the length of the time interval.

  For example, in order to reduce the transfer bit rate of the sub-content (sub-transmission signal) provided from the providing device 11, the audio synchronization feature amount included in the sub-transmission signal is set to a low rate, while the microphone 84 is set. The feature amount for voice synchronization calculated from the voice picked up in step 1 may be set at a high rate because there is no need for transmission.

  In such a case, for example, as shown in FIG. 7, it is conceivable to downsample the audio synchronization feature amount calculated from the sound collected by the microphone 84 using the same method as the downsampling unit 56. In FIG. 7, the vertical axis of the peak information as the voice synchronization feature amount indicated by the arrows Q11 to Q14 indicates the time delay τ, and the horizontal axis indicates the time interval i. One square represents peak information in one time section.

  In this example, on the providing apparatus 11 side, after peak information as a feature amount for voice synchronization is obtained as indicated by an arrow Q11, the peak information is downsampled, and a peak with a longer time interval indicated by an arrow Q12 is obtained. Information is transmitted to the sub-receiving device 85. Here, the peak information of the time interval of 8 msec is frame rate converted (downsampled) into the peak information of the time interval of 32 msec.

  On the other hand, the audio synchronization feature value calculation unit 121 of the sub reception device 85 calculates the audio synchronization feature value from the audio signal obtained by collecting the audio of the main content reproduced by the main reception device 81, and as a result. The peak information indicated by the arrow Q13 is obtained as the voice synchronization feature amount. Here, the peak information indicated by the arrow Q13 is calculated for each 8 msec time interval.

  Thus, the length of the time interval, that is, the frame rate, differs between the voice synchronization feature value obtained by the voice synchronization feature value calculation unit 121 and the voice synchronization feature value received from the providing apparatus 11. Therefore, the frame rate conversion unit 181 downgrades the frame rate conversion for the audio synchronization feature quantity obtained by the audio synchronization feature quantity calculation unit 121 so that the frame rates of the audio synchronization feature quantities match. A sample is performed to obtain peak information as a voice synchronization feature amount indicated by an arrow Q14. The voice synchronization feature amount indicated by the arrow Q14 is peak information of a time interval of 32 msec.

  After the frame rate (the length of the time interval) is aligned in this way, the synchronization calculation is performed using the audio synchronization feature amount. Thus, by performing the downsampling of the audio synchronization feature quantity on the sub-receiving device 85 side, it is possible to cope with an arbitrary frame rate (bit rate).

  In addition, the voice synchronization feature value transmitted to the sub receiver 85 has a high rate, but the voice synchronization feature value calculated from the sound collected by the microphone 84 may have a low rate. For example, there are cases where the computing resources of the sub receiving device 85 are not abundant and the frame shift amount is increased in order to reduce the amount of computation necessary for calculating the voice synchronization feature value.

  In such a case, for example, the frame rate of the voice synchronization feature amount included in the sub-transmission signal indicated by the arrow Q21 in FIG. Thus, the voice synchronization feature amount indicated by the arrow Q22 is obtained. In FIG. 8, the vertical axis of the peak information as the voice synchronization feature value indicated by the arrows Q21 to Q23 indicates the time delay τ, and the horizontal axis indicates the time interval i. One square represents peak information in one time section.

  In this example, the peak information of the 8 msec time interval is frame rate converted (downsampled) to the peak information of the 32 msec time interval.

  In addition, the audio synchronization feature value calculation unit 121 of the sub reception device 85 calculates the audio synchronization feature value from the audio signal obtained by collecting the audio of the main content reproduced by the main reception device 81, and as a result. The peak information indicated by the arrow Q23 is obtained as the voice synchronization feature amount. Here, the peak information indicated by the arrow Q23 is calculated for each 32 msec time interval.

  In this way, the audio synchronization feature amount included in the sub transmission signal is down-sampled, and the frame rate of the audio synchronization feature amount included in the sub transmission signal and the audio synchronization calculated by the sub reception device 85 are obtained. The frame rate of the feature amount may be matched.

  Furthermore, in the above description, the frame rate is matched by down-sampling a voice synchronization feature with a higher frame rate. However, the frame rate is matched by up-sampling a voice synchronization feature with a lower frame rate. May be.

  In such a case, for example, as shown in FIG. 9, the frame rate of the audio synchronization feature amount included in the sub transmission signal is upsampled by the frame rate conversion unit 182. In FIG. 9, the vertical axis of the peak information as the voice synchronization feature value indicated by the arrows Q31 to Q34 indicates the time delay τ, and the horizontal axis indicates the time interval i. One square represents peak information in one time section.

  In this example, on the providing device 11 side, after peak information as a feature amount for voice synchronization is obtained as indicated by an arrow Q31, downsampling is performed, and peak information having a longer time interval is indicated by an arrow Q32. And transmitted to the sub-receiving device 85. Here, the peak information of the time interval of 8 msec is frame rate converted (downsampled) into the peak information of the time interval of 32 msec.

  On the other hand, the audio synchronization feature value calculation unit 121 of the sub reception device 85 calculates the audio synchronization feature value from the audio signal obtained by collecting the audio of the main content reproduced by the main reception device 81, and as a result. The peak information indicated by the arrow Q33 is obtained as the voice synchronization feature amount. Here, the peak information indicated by the arrow Q33 is calculated for each time interval of 8 msec.

  In this example, the frame rate does not match between the voice synchronization feature quantity calculated by the voice synchronization feature quantity calculation unit 121 and the voice synchronization feature quantity received from the providing apparatus 11.

  Therefore, the frame rate conversion unit 182 upsamples the peak information as the voice synchronization feature value received from the providing apparatus 11 and calculates the peak information of the 8 msec time section indicated by the arrow Q34 to be used for the synchronization calculation. Align the time synchronization granularity of the features for voice synchronization. For example, the frame rate conversion unit 182 upsamples the peak information by calculating the following equation (5).

In the calculation of Expression (5), the time delay τ is the same, and the pre-upsample 1 is in the same positional relationship as the four peak information P _i (τ) after the upsample adjacent in the time direction (time interval direction). The values of the two pieces of peak information P ′ _i (τ) are the values of the four pieces of peak information P _i (τ) after the upsampling as they are.

  As described above, by appropriately upsampling the voice synchronization feature amount used for the synchronization calculation in accordance with a higher frame rate, it is possible to realize a pseudo-high resolution synchronization accuracy.

  Further, in order to reduce computation resources in the sub receiving device 85, both the voice synchronization feature value included in the sub transmission signal and the voice synchronization feature value calculated by the voice synchronization feature value calculation unit 121 are obtained. It is also possible to downsample.

  By having the frame rate conversion unit 181 and the frame rate conversion unit 182 as described above, it is possible to perform synchronization between audio synchronization feature quantities having different frame rates. In addition, various frame rates can be specified according to computing resources, transmission bands, etc., and the flexibility of the system can be enhanced.

  Returning to the description of FIG. 6, the block integration unit 183 receives the time-series data of the audio synchronization feature amount for each time interval of the main content from the frame rate conversion unit 181 and receives a plurality of continuous (for example, 64) time intervals. Are integrated in units of blocks. The block integration unit 183 supplies the time series data of the feature amount for voice synchronization in units of blocks to the similarity calculation unit 185.

  The block integration unit 184 receives the time-series data of the audio synchronization feature amount for each time interval of the main content from the frame rate conversion unit 182, and sets a plurality of (for example, 64) time intervals as one block. Integrate with. The block integration unit 184 supplies the time series data of the feature amount for voice synchronization in units of blocks to the similarity calculation unit 185.

  Note that the plurality of time intervals constituting the block may not be continuous. For example, a plurality of even-numbered time intervals can be one block, or a plurality of odd-numbered time intervals can be one block. In this case, since the thinning process can be performed on the time-series data of the voice synchronization feature value for each time interval, the amount of calculation can be reduced.

  The similarity calculation unit 185 calculates the similarity between the time-series data of the feature quantities for voice synchronization in units of blocks supplied from the block integration unit 183 and the block integration unit 184, and represents the similarity between the blocks. Generate a similarity matrix. The similarity calculation unit 185 supplies the similarity matrix to the optimum path search unit 186.

  The optimal path search unit 186 searches the similarity matrix supplied from the similarity calculation unit 185 for a path with the optimal similarity, and uses information representing the time difference between the two blocks corresponding to the similarity on the path as a voice feature. It is generated as synchronization correction information based on the amount. Then, the optimum path search unit 186 supplies synchronization correction information based on the audio feature amount to the reproduction processing unit 127.

  As described above, the synchronization calculation unit 126 generates synchronization correction information based on the audio feature amount based on the pitch information of the audio signal. Therefore, even when different noise is included in each audio signal, the synchronization correction information based on the audio feature amount can be generated robustly.

  That is, when a human hears a plurality of sounds having frequency characteristics, noise sounds are included by perceiving sounds having the same fundamental frequency, that is, sounds having the same pitch, as a common component Even so, the common component can be easily perceived. In consideration of this, the present technology generates synchronization correction information based on pitch information, thereby generating synchronization correction information robust to noise sounds.

<Similarity calculation and search for optimal similarity path>
Here, the calculation of the similarity and the search for the path with the optimum similarity will be described.

  FIG. 10 is a diagram illustrating a block that is a target of similarity calculation.

  In FIG. 10, i is the index of the block of the feature value for voice synchronization obtained by the feature amount calculation unit 121 for voice synchronization, and j is the block of the feature amount for voice synchronization included in the sub received signal. Is the index of More specifically, these audio synchronization feature quantities are appropriately frame rate converted by the frame rate conversion unit 181 and the frame rate conversion unit 182, but here, for the sake of simplicity, the frame rate conversion is performed. The description of the similarity calculation will be continued as it is not performed.

  X (i) represents time-sequential data of the voice synchronization feature quantity of the block of index i among the voice synchronization feature quantities obtained by the voice synchronization feature quantity calculation unit 121, and Y (j) represents , Represents the time-sequential data of the voice synchronization feature quantity of the index j block among the voice synchronization feature quantities included in the sub received signal.

  As shown in FIG. 10, the target of similarity calculation is n × m combinations of each of n X (i) and m Y (j).

  FIG. 11 is a diagram illustrating a method for calculating similarity.

  In the matrix of FIG. 11, the horizontal axis represents the intra-block time interval number representing the number from the beginning of the time interval within the block, and the vertical axis represents the index τ. A white square indicates that the time series data P (τ) of the voice synchronization feature amount of the index τ of the time section of the corresponding intra-block time section number is 0, and the black square indicates the time series data. It represents that P (τ) is 1. Furthermore, in the example of FIG. 11, it is assumed that the number of time intervals constituting the block is four and τ is 0 to 3.

  As shown in FIG. 11, when calculating the similarity between X (i) and Y (j), first, the logical product X (i) ∩Y (j) of X (i) and Y (j) is calculated. Next, the logical sum X (i) ∪Y (j) of X (i) and Y (j) is calculated. For example, as shown in FIG. 11, when calculating the similarity between X (i) and Y (j) consisting of 9 0s and 7 1s, first, it consists of 12 0s and 4 1s. The logical product X (i) ∩Y (j) is calculated, and the logical sum X (i) ∪Y (j) consisting of six 0s and ten 1s is calculated.

  Then, according to the following equation (6), Number (X (i) ∩Y (j)) which is the number of 1 of the logical product X (i) ∩Y (j) and Number (X which is the number of 1 of the logical sum) Based on (i) ∪Y (j)), the similarity A (i, j) between X (i) and Y (j) is calculated.

  In the example of FIG. 11, since Number (X (i) ∩Y (j)) is 4 and Number (X (i) ∪Y (j)) is 10, the similarity A (i, j) is 0.4.

  In addition, when the sum S (τ) of periodicity information is used as the time series data of the feature amount for voice synchronization, a method of calculating similarity using a cosine distance is used as a method of calculating similarity. be able to.

  The similarity matrix is information indicating the similarity A (i, j) of each point corresponding to the index i and the index j, for example, where the horizontal axis is the index j and the vertical axis is the index i.

  The optimal path search unit 186 searches for a path having the maximum integrated value of the path similarity on the similarity matrix as a path with the optimal similarity using dynamic programming. The optimal path search unit 186 generates an index difference ij corresponding to the similarity on the path having the optimal similarity as synchronization correction information based on the audio feature amount.

<Description of transmission processing>
Next, the operation of the providing device 11 will be described.

  When the main channel signal and the sub channel signal that are time-synchronized with each other are supplied, the providing device 11 performs a transmission process and transmits the main transmission signal and the sub transmission signal. Hereinafter, the transmission process by the providing apparatus 11 will be described with reference to the flowchart of FIG.

  In step S11, the voice synchronization feature value calculation unit 23 performs a voice synchronization feature value calculation process, calculates a voice synchronization feature value from the voice signal constituting the supplied main channel signal, and performs a multiplexing process. To the unit 24.

  Details of the voice synchronization feature value calculation processing will be described later.

  In step S12, the conversion unit 21 generates a main transmission signal by converting the supplied main channel signal into a signal of a predetermined transmission format defined by the system, and the obtained main transmission signal is output to the output unit 22. Supply.

  In step S <b> 13, the output unit 22 transmits the main transmission signal supplied from the conversion unit 21.

  In step S <b> 14, the multiplexing processing unit 24 multiplexes the audio synchronization feature quantity and the subchannel signal, and supplies the sub-transmission signal obtained as a result to the output unit 25.

  For example, the multiplexing processing unit 24 uses the supplied main channel signal so that the time synchronization relationship between the audio synchronization feature value from the audio synchronization feature value calculation unit 23 and the supplied subchannel signal matches. In addition, the audio synchronization feature quantity and the subchannel signal are multiplexed according to a transmission format defined by the system.

  Thereby, for example, the sub transmission signal shown in FIG. 13 is obtained.

  In the example of FIG. 13, the section T11 and the section T12 in the bit stream as the sub transmission signal each include an image signal, an audio signal, and an audio synchronization feature amount for one frame.

  For example, the image signal and the audio signal included in the section T11 are sub-channel signals for one frame, and the audio synchronization feature amount included in the section T11 is the main channel of the frame that temporally corresponds to the sub-channel signal. This is a voice synchronization feature amount extracted from the signal. As described above, in the sub transmission signal, the sub-channel signal of the same frame and the audio synchronization feature amount are associated and multiplexed, and on the receiving side of the sub-transmission signal, the sub-transmission signal is associated with the sub-channel signal of each frame. The specified feature amount for voice synchronization can be specified.

  Returning to the description of the flowchart of FIG. 12, in step S15, the output unit 25 transmits the sub transmission signal supplied from the multiplexing processing unit 24, and the transmission process ends.

  As described above, the providing apparatus 11 generates a sub transmission signal by associating and multiplexing the audio synchronization feature amount obtained from the main channel signal and the sub channel signal, and generates the sub transmission signal and the main transmission signal. Send a send signal.

  By transmitting the audio synchronization feature quantity in association with the subchannel signal in this way, on the receiving side, when the main channel signal and the subchannel signal are received by a plurality of different devices via different transmission paths. However, the main content and the sub-content can be reproduced in synchronization using the audio synchronization feature amount.

<Explanation of feature calculation processing for voice synchronization>
Next, with reference to the flowchart of FIG. 14, the audio synchronization feature quantity calculation process corresponding to the process of step S11 of FIG. 12 will be described.

  In step S41, the frequency band dividing unit 51 divides the supplied audio signal into time intervals of about several tens msec to 100 msec using a window function.

  In step S42, the frequency band dividing unit 51 divides the audio signal for each time interval into four frequency bands using a plurality of bandpass filters. The frequency band dividing unit 51 supplies the audio signal of each frequency band to each of the periodicity detecting unit 52-1 to the periodicity detecting unit 52-4.

  In step S43, the periodicity detection unit 52 calculates the autocorrelation function x (b, τ) of the audio signal for each time interval of the predetermined frequency band supplied from the frequency band dividing unit 51, thereby The periodicity information is extracted and supplied to the periodicity intensity detection unit 53 and the periodicity information integration unit 54. The process of step S43 is performed for each periodicity detection unit 52.

  In step S <b> 44, the periodic strength detector 53 calculates the periodic strength for each time interval based on the periodicity information for each time interval supplied from the periodicity detector 52. And the periodicity intensity | strength detection part 53 produces | generates the periodicity intensity | strength information for every time section by binarizing the intensity | strength of the periodicity for every time section by whether it exceeded the threshold value, and the periodicity information integration part 54. In addition, the process of step S44 is performed for each periodic strength detector 53.

  In step S <b> 45, the periodicity information integration unit 54 converts the periodicity information for each time interval supplied from the periodicity detection unit 52 and the periodicity information for each time interval supplied from the periodicity intensity detection unit 53. Based on the above formula (1), periodic integration processing is performed. The periodicity information integration unit 54 supplies the peak detection unit 55 with the sum S (τ) of periodicity information for each time interval obtained as a result of the periodicity integration processing.

  In step S46, the peak detection unit 55 performs peak detection on the sum S (τ) of the periodicity information supplied from the periodicity information integration unit 54 for each time interval, and generates peak information P (τ). And supplied to the down sample unit 56.

  In step S47, the downsampling unit 56 performs peak information downsampling processing by integrating the peak information P (τ) in a plurality of time intervals supplied from the peak detection unit 55 into one time interval.

  The down-sampling unit 56 supplies the peak information for each time interval obtained in this way to the multiplexing processing unit 24 as time-sequential data of the audio synchronization feature value for each time interval, and calculates the audio synchronization feature value. The process ends. When the voice synchronization feature value calculation process is completed, the process proceeds to step S12 in FIG.

  Since the voice synchronization feature value calculating unit 23 calculates the voice synchronization feature value based on the periodicity information as described above, the voice synchronization feature value can be generated robustly.

<Description of main content playback processing>
When the main transmission signal is transmitted from the providing device 11, the content reproduction system acquires the main transmission signal as the main reception signal and reproduces the main content. Hereinafter, the main content reproduction process by the content reproduction system will be described with reference to the flowchart of FIG.

  In step S <b> 71, the input unit 111 acquires the main reception signal and supplies it to the reproduction processing unit 112. For example, the input unit 111 acquires the main reception signal by receiving the main reception signal transmitted from the providing apparatus 11.

  In step S72, the reproduction processing unit 112 reproduces the main content based on the main reception signal supplied from the input unit 111, and the main content reproduction process ends.

  For example, the reproduction processing unit 112 extracts the image signal and audio signal of the main content from the main reception signal, supplies the image signal to the display unit 82 for reproduction, and supplies the audio signal to the speaker 83 for reproduction. Thereby, the main content is reproduced.

  As described above, the content playback system acquires the main reception signal and plays back the main content.

<Description of sub-content playback processing>
Further, in synchronization with the reproduction of the main content, the content reproduction system acquires the sub received signal and reproduces the sub content. Hereinafter, with reference to the flowchart of FIG. 16, sub-content reproduction processing by the content reproduction system will be described.

  In step S <b> 101, the input unit 123 acquires the sub reception signal and supplies it to the separation processing unit 124. For example, the input unit 123 acquires the sub reception signal by receiving the sub transmission signal transmitted from the providing apparatus 11 as the sub reception signal.

  In step S102, the separation processing unit 124 separates the sub reception signal supplied from the input unit 123 into the subchannel signal and the audio synchronization feature quantity, and buffers the separated subchannel signal and the audio synchronization feature quantity. 125 to be recorded.

  In step S <b> 103, the microphone 84 collects the audio of the main content output from the speaker 83, and supplies the audio signal obtained as a result to the audio synchronization feature quantity calculation unit 121. For example, in step S103, the sound of the main content reproduced by the process in step S72 in FIG. 15 is collected.

  In step S104, the audio synchronization feature value calculation unit 121 performs audio synchronization feature value calculation processing, calculates the audio synchronization feature value from the audio signal supplied from the microphone 84, supplies it to the buffer 122, and records it. Let

  Note that, as the voice synchronization feature amount calculation processing, the processing of steps S131 to S136 shown in the flowchart of FIG. 17 is performed, and these processing are the same as the processing of steps S41 to S46 of FIG. Description is omitted. However, in the audio synchronization feature value calculation process shown in FIG. 17, the audio synchronization feature value is calculated from the audio signal supplied from the microphone 84 and stored in the buffer 122. Further, in the voice synchronization feature quantity calculation unit 121, the peak information obtained by the peak detection unit 155 is used as the voice synchronization feature quantity.

  Returning to the description of the flowchart of FIG. 16, in step S <b> 105, the synchronization calculation unit 126 performs synchronization correction information generation processing, generates synchronization correction information based on the audio feature amount, and supplies the synchronization correction information to the reproduction processing unit 127. Although details of the synchronization correction information generation process will be described later, in this process, the audio synchronization feature quantity recorded in the buffer 122 is compared with the audio synchronization feature quantity recorded in the buffer 125. The synchronization correction information based on the audio feature amount for synchronizing the main content and the sub-content is generated.

  In step S106, the reproduction processing unit 127 corrects the reproduction timing of the subchannel signal recorded in the buffer 125 based on the synchronization correction information based on the audio feature amount supplied from the synchronization calculation unit 126, and performs the correction. The sub contents are reproduced based on the sub channel signal.

  That is, the reproduction processing unit 127 supplies the image signal and audio signal constituting the subchannel signal to the display unit 86 and the speaker 87 for reproduction or reproduction later or earlier by the time indicated by the synchronization correction information based on the audio feature amount. In other words, the sub-content part at the reproduction time corresponding to the main content part reproduced at the current time, which is specified from the synchronization correction information based on the audio feature amount, is reproduced.

  For example, the adjustment (correction) of the reproduction position for synchronizing the sub content with the main content is performed in a silent section of the sub content or the main content.

  The display unit 86 displays the sub-content image based on the image signal supplied from the reproduction processing unit 127, and the speaker 87 outputs the sub-content audio based on the audio signal supplied from the reproduction processing unit 127. Output.

  In this way, when the sub-content is played back in synchronization with the main content, the sub-content playback processing ends.

  As described above, the content reproduction system calculates the audio synchronization feature amount from the audio signal obtained by collecting the audio of the main content being reproduced, and obtains the audio synchronization feature amount and the sub-reception. Synchronization correction information based on the voice feature amount is calculated using the voice synchronization feature amount included in the signal. Further, the content reproduction system reproduces the sub-content in synchronization with the main content using the obtained synchronization correction information.

  As described above, the synchronization correction information based on the audio feature amount is obtained using the audio synchronization feature amount extracted from the audio signal obtained by collecting the sound and the audio synchronization feature amount included in the sub reception signal. By calculating, even if the transmission paths of the main content and the sub-content are different, those contents can be reproduced in synchronization.

  In this example, the synchronization calculation of the voice synchronization feature value, that is, the matching process is performed every frame, but the synchronization calculation of the voice synchronization feature value is not necessarily performed continuously in time, and is intermittent. May be performed automatically. However, when the synchronization calculation is continuously performed, correction can be performed without a sense of incompatibility when correcting the playback time (playback position) of the sub-content.

<Description of synchronization correction information generation processing>
Further, the synchronization correction information generation process corresponding to the process of step S105 of FIG. 16 will be described with reference to the flowchart of FIG.

  In step S161, the frame rate conversion unit 181 and the frame rate conversion unit 182 perform frame rate conversion processing as necessary.

  That is, the frame rate conversion unit 181 reads the time series data of the audio synchronization feature value for each time interval of the main content from the buffer 122, performs frame rate conversion, that is, down-samples the audio synchronization feature value as necessary, This is supplied to the block integration unit 183. Also, the frame rate conversion unit 182 reads the time series data of the audio synchronization feature value for each time interval of the main content from the buffer 125, and converts the audio synchronization feature value to frame rate conversion, that is, down-sample or up, as necessary. Sample and supply to the block integration unit 184.

  In step S162, the block integration unit 183 and the block integration unit 184 integrate the time series data of the audio synchronization feature amount.

  Specifically, the block integration unit 183 receives the time-series data of the feature amount for audio synchronization for each time interval of the main content from the frame rate conversion unit 181. Then, the block integration unit 183 integrates the time series data of the feature quantities for voice synchronization for each supplied time section in units of blocks, with a plurality of continuous (for example, 64) time sections as one block. It supplies to the calculation part 185.

  Further, the block integration unit 184 receives supply of time-series data of feature amounts for audio synchronization for each time interval of the main content from the frame rate conversion unit 182. Then, the block integration unit 184 integrates the time series data of the feature amount for voice synchronization for each supplied time interval in units of blocks, with a plurality of continuous (for example, 64) time intervals as one block. It supplies to the calculation part 185.

  In step S163, the similarity calculation unit 185 calculates the similarity between the time series data of the feature quantities for voice synchronization in units of blocks supplied from the block integration unit 183 and the block integration unit 184. A similarity matrix representing the similarity is generated. The similarity calculation unit 185 supplies the similarity matrix to the optimum path search unit 186.

  In step S164, the optimal path search unit 186 searches for a path with the optimal similarity from the similarity matrix supplied from the similarity calculation unit 185, and generates synchronization correction information based on the audio feature amount. Then, the optimum path search unit 186 supplies synchronization correction information based on the audio feature amount to the reproduction processing unit 127, and the synchronization correction information generation process ends.

  As described above, the content reproduction system generates the synchronization correction information based on the audio feature amount based on the periodicity information, so that the synchronization correction information can be generated robustly.

  In addition, although the case where there is one main content has been described above, there may be a plurality of main contents.

  In such a case, the audio synchronization feature value calculating unit 23 of the providing apparatus 11 calculates the audio synchronization feature value for each of the plurality of main contents, and the multiplexing processing unit 24 includes a subchannel signal of one subcontent, Multiple feature values for audio synchronization of main contents are multiplexed and used as a sub transmission signal. The output unit 22 transmits a main transmission signal obtained from the main channel signals of a plurality of main contents.

  Furthermore, in this case, in the content reproduction system shown in FIG. 4, the reproduction processing unit 112 selects and reproduces one of the plurality of main contents. Further, the input unit 123 receives a sub reception signal in which a plurality of main content audio synchronization feature quantities are associated with one sub channel signal.

  Then, the synchronization calculation unit 126 compares the audio synchronization feature quantity of each main content acquired by the input unit 123 with the audio synchronization feature quantity obtained by the audio synchronization feature quantity calculation unit 121 to obtain a similarity. The main content being calculated by the speaker 83 is specified. For example, as a result of matching with the audio synchronization feature value, the main content of the audio synchronization feature value having the highest similarity is assumed to be the main content being reproduced.

  When the main content being reproduced is specified, the reproduction position of the sub-content is corrected based on the synchronization correction information obtained for the audio synchronization feature amount of the specified main content. That is, the synchronization calculation unit 126 generates synchronization correction information based on the audio feature amount for synchronizing the identified main content and sub-content.

<Application example 1 of this technology>
In addition, the present technology described above can be applied to various types of systems.

  For example, the present technology can be applied to the system shown in FIG.

  In the system shown in FIG. 19, a providing device 211 such as a broadcasting station corresponds to the providing device 11 in FIG. The providing device 211 provides main content and sub-content.

  In this example, the providing apparatus 211 transmits the main transmission signal to the main receiving device 212 by broadcasting the main transmission signal of the main content by, for example, a broadcast wave. Then, the main reception device 212 receives the main transmission signal transmitted by the broadcast wave as the main reception signal and reproduces the main content. At this time, the main receiving device 212 outputs the sound of the main content from the speaker 213 provided in the main receiving device 212.

  Therefore, in this example, the main receiving device 212 is configured by the main receiving device 81, the display unit 82, and the speaker 83 shown in FIG. In this case, the input unit 111 receives a main reception signal broadcast by a broadcast wave. Further, the speaker 213 corresponds to the speaker 83 of FIG.

  For example, the main receiving device 212 is a television receiver or the like, and the user views main content reproduced by the main receiving device 212.

  On the other hand, the providing apparatus 211 also transmits a sub transmission signal. In this example, the providing device 211 transmits the sub transmission signal to the sub reception device 215 by streaming distribution or the like via the communication network 214 such as the Internet. Here, the sub transmission signal is transmitted by so-called push type communication.

  The sub reception device 215 includes, for example, a tablet-type terminal device and the like, receives a sub transmission signal transmitted via the communication network 214 as a sub reception signal, and reproduces the sub content. That is, the sub reception device 215 displays the sub content image on the built-in display unit and outputs the sub content sound from the built-in speaker.

  At this time, the sub reception device 215 collects the audio of the main content output from the speaker 213 and calculates the audio synchronization feature quantity, and is included in the obtained audio synchronization feature quantity and the sub reception signal. Synchronization correction information based on the voice feature is generated using the voice synchronization feature. Then, the sub receiving device 215 uses the synchronization correction information based on the audio feature amount to reproduce the sub content in synchronization with the main content.

  As a result, the main content played back by the main receiving device 212 and the sub-content played back by the sub-receiving device 215 are played back in synchronization with each other, and the user appropriately watches the main content while watching and listening to the sub-contents. Can be watched. That is, the main content can be enjoyed while using the sub-content as auxiliary information for the main content, for example.

  In this example, the sub-channel signal of the sub-content is, for example, an image signal of a video at a different angle from the video of the main content, an audio signal of commentary audio for the main content, character information related to the main content, or the like.

  In this case, the sub reception device 215 includes, for example, the microphone 84, the sub reception device 85, the display unit 86, and the speaker 87 illustrated in FIG. Therefore, the input unit 123 receives a sub transmission signal transmitted via the communication network 214 as a sub reception signal.

  As described above, in the example of FIG. 19, the main content and the sub-content transmitted through different transmission paths can be easily and accurately reproduced on the receiving side in push-type communication. In this example, the sub transmission signal needs to be transmitted prior to the main transmission signal. That is, the transmission of the main transmission signal and the sub transmission signal is performed with a time difference taking into account the difference (arrival time difference) between the arrival time of the main transmission signal at the main reception device 212 and the arrival time of the sub transmission signal at the sub reception device 215. Need to be done.

<Application example 2 of this technology>
Moreover, this technique is applicable also to the system shown, for example in FIG. In FIG. 20, the same reference numerals are given to the portions corresponding to those in FIG. 19, and the description thereof is omitted.

  In the example of FIG. 20, as in the example of FIG. 19, the main transmission signal is transmitted from the providing device 211 to the main reception device 212 by broadcast waves, that is, by push-type communication.

  On the other hand, the sub transmission signal is transmitted from the server 241 to the sub reception device 215 via the communication network 214. Note that the server 241 acquires and records the sub-transmission signal from the providing device 211 or the like in advance by some method.

  In this example, the sub transmission signal is transmitted by so-called pull-type communication. Therefore, the server 241 transmits the sub transmission signal to the sub reception device 215 via the communication network 214 when the sub reception device 215 requests transmission of the sub transmission signal.

  That is, the input unit 123 of the sub-receiving device 85 in FIG. 4 corresponding to the sub-receiving device 215 transmits a sub-transmission signal transmission request to the server 241 and transmits the sub-transmission signal transmitted from the server 241 in response to the transmission request. The transmission signal is received as a sub reception signal.

  In this case, the sub reception device 215 can receive and record the sub transmission signal in advance before broadcasting the main content. Therefore, if the sub transmission signal is received and recorded in advance, it is possible to prevent a situation in which the sub content cannot be reproduced in synchronization with the main content due to the state of the communication network 214 or the like when the main content is broadcast. .

  When the main reception device 212 starts to reproduce the main content, the sub reception device 215 collects the audio of the main content output from the speaker 213 and calculates the audio synchronization feature amount. Then, the sub reception device 215 generates synchronization correction information based on the audio feature amount using the obtained audio synchronization feature amount and the audio synchronization feature amount included in the sub reception signal, and the synchronization correction information Is used to reproduce the sub-content in synchronization with the main content.

  As described above, in the example of FIG. 20, the sub reception device 215 can acquire the sub reception signal at a timing convenient for itself.

<Application example 3 of this technology>
Moreover, this technique is applicable also to the system shown, for example in FIG. In FIG. 21, the same reference numerals are given to the portions corresponding to those in FIG. 20, and the description thereof is omitted.

  In the example of FIG. 21, the main content, that is, the main transmission signal is provided by a server 271 different from the server 241. That is, the server 271 transmits the recorded main transmission signal to the main receiving device 212 via the communication network 272 when requested by the main receiving device 212. That is, in this example, the main transmission signal is transmitted by pull-type communication.

  Specifically, the input unit 111 of the main receiving device 81 in FIG. 4 corresponding to the main receiving device 212 transmits a transmission request for the main transmission signal to the server 271 and is transmitted from the server 271 in response to the transmission request. The main transmission signal received is received as the main reception signal.

  In this case, the main receiving device 212 can receive and record the main transmission signal in advance. Therefore, if the main transmission signal is received and recorded in advance, it is possible to prevent a situation in which the reproduction of the main content is interrupted or stopped during the reproduction of the main content due to the state of the communication network 272 or the like.

  Further, the sub transmission signal is transmitted by the server 241 by pull-type communication as in the example of FIG.

  When the main reception device 212 starts to reproduce the main content, the sub reception device 215 collects the audio of the main content output from the speaker 213 and calculates the audio synchronization feature amount. Then, the sub reception device 215 generates synchronization correction information based on the audio feature amount using the obtained audio synchronization feature amount and the audio synchronization feature amount included in the sub reception signal, and the synchronization correction information Is used to reproduce the sub-content in synchronization with the main content.

  In this way, in the example of FIG. 21, the main reception device 212 and the sub reception device 215 can acquire the main reception signal and the sub reception signal at a timing convenient for themselves.

  Even if the communication network 272 is the same communication network as the communication network 214, if the transmission timing of the main transmission signal and the sub-transmission signal, the receiving device, etc. are different, these main transmission signal and sub-transmission are usually used. The signal transmission path is different.

<Second Embodiment>
<Features of this technology>
By the way, as in the examples shown in the above (A1) to (A4), a plurality of media contents having a time synchronization relationship are received by a plurality of devices through broadcasting, an IP (Internet Protocol) network, and the received media contents are received. An application program that plays in synchronization is assumed.

  In order to realize such functionality, content is delivered by broadcasting like Hybridcast, and at the same time, additional content is delivered individually by the IP network. The content receiver uses the content delivered by broadcasting and the IP network. Research and development of a system for a broadcasting / communication cooperation service in which additional content delivered is synchronized in time and output simultaneously.

  For example, “Hyoji Matsumura, Yoshiaki Shiga, Michael J Evans,“ Examination of a personalized system for broadcasting programs linked with Internet distribution information ”, Proceedings of Annual Conference of the Institute of Image Information and Television Engineers, August 26, 2009 , P. 3-8 "(hereinafter also referred to as non-patent document 1).

  Also, “Japan Broadcasting Corporation,“ HybridcastTM Overview and Technology ”, NHK STRL R & D, no.124, p.10-17, November 2010, Japan Broadcasting Publishing Association, http://www.nhk.or.jp /strl/publica/rd/rd124/PDF/P10-17.pdf (hereinafter also referred to as Non-Patent Document 2), “Japan Broadcasting Corporation,“ Technology supporting HybridcastTM ”, NHK STRL R & D, no.133, p. .20-27, May 2012, Japan Broadcasting Publishing Association, http://www.nhk.or.jp/strl/publica/rd/rd133/PDF/P20-27.pdf ”(hereinafter referred to as Non-Patent Document 3) (Also referred to as “Hybridcast”).

  With Hybridcast, additional content with presentation time information (PTS (Presentation Time Stamp)) based on the reference clock (PCR (Program Clock Reference)) of the broadcast stream is streamed at the same time as the broadcast content is sent, or a little earlier. The basic principle is that the receiver has a sufficient amount of buffer to absorb delays and fluctuations in the communication content, delays the broadcast content, and synchronizes both by comparing the time stamps.

  For example, according to Non-Patent Document 2, it has been confirmed that synchronization can be achieved with an accuracy of about one video frame (33 ms) in a prototype environment in which both receivers are in the same device.

  The device that receives the additional content may be a device that is independent of the broadcast content receiver, such as a smart phone or a tablet personal computer that is wirelessly connected to the IP network. In such a case, the broadcast content receiver needs to provide presentation time information (time stamp) to a device that receives the additional content. This is usually linked via an IP network.

  In addition, even if it is not broadcast, multiple contents are distributed only via a network such as the IP network, a time stamp is added using Coordinated Universal Time (UTC) as a reference clock, and synchronization is performed on the receiver side. Realization of the output system can be easily imagined.

  In fact, when the above-mentioned broadcasting / communication cooperation service is used by an independent receiver, it is difficult for the method based on the time stamp comparison to achieve strict synchronization due to the following two factors.

  First, since the broadcast content receiver and the additional content receiver are independent electronic devices, there is a difference in the system clock, and a synchronization shift occurs as time passes.

  Second, the user is at a certain distance from a broadcast content receiver such as a television receiver, has an additional content receiver such as a smartphone or tablet personal computer, and is distributed via the IP network. A usage pattern of enjoying content is assumed. When the audio content is included in the broadcast content and the additional content in this usage pattern, it becomes difficult to achieve exact synchronization at the viewing position of the user.

  For example, when the user is 10 m away from the broadcast content receiver, 10 (m) / 340 (m / s) = about 30 (ms) for the audio signal output from the broadcast content receiver to reach the user position It will take time. Here, the sound speed is about 340 (m / s).

  In the first embodiment, the additional content receiver picks up the audio output from the broadcast content receiver, calculates the audio synchronization feature, and synchronizes the audio of the broadcast content distributed over the IP network. This is a technique for performing synchronous calculation with feature quantities. However, when the transmission delay or fluctuation of the IP network is large, it is necessary to search the synchronization position over a wide range, and the processing amount increases.

  Therefore, by providing the above-described providing apparatus and content reproduction system with the following features B11 to B20, a plurality of contents acquired through different paths can be synchronized with a smaller amount of processing. .

(Feature B11)
The media content is a data stream in which video, audio, image, character information, etc. are multiplexed.

  As for the transmission of the data stream of (feature B11), the transmission of media content in a network such as a broadcast wave or the Internet is assumed, and the logical transmission path occupied by the multiplexed data stream is referred to as a transmission path.

(Feature B12)
Multiple media contents to be transmitted have a time synchronization relationship.

(Feature B13)
At least one of a plurality of media contents to be transmitted is defined as a main channel signal, and each remaining media content is defined as a subchannel signal.

(Feature B14)
Presentation time information (PTC) is generated for each of the main channel signal and the sub-channel signal from the reference time signal.

  Here, the reference clock signal is a broadcast stream reference clock (PCR) or Coordinated Universal Time (UTC).

(Feature B15)
The presentation time information of the main channel signal is multiplexed with the main channel signal, and a main transmission signal is generated and transmitted. On the other hand, a feature value for voice synchronization is also calculated from the voice signal of the main channel signal.

(Feature B16)
The time synchronization relationship between the main channel signal and the subchannel signal is matched, and the presentation time information of the subchannel signal, the audio synchronization feature quantity of the main channel signal, and the subchannel signal multiplexing processing according to the transmission format specified by the system To generate a sub transmission signal.

(Feature B17)
The main receiving device acquires and separates the main reception signal, and outputs sound based on the audio signal through a speaker or the like when reproducing the main channel signal. At the same time, the main receiving device presents the presentation time information of the received main channel signal so that it can be referred to from outside or acquired.

  For example, the presentation time information of the main channel signal is provided by a software API (Application Programming Interface) so that it can be referred from the outside via an IP network connection by wireless communication.

(Feature B18)
The sub reception device acquires and separates the sub reception signal, compares the presentation time information of the received sub channel signal with the presentation time information of the main channel signal acquired from the main reception device, and synchronizes correction information based on the presentation time information Is generated.

(Feature B19)
The sub receiving device collects the sound of the main channel signal output from the speaker by the main receiving device using a microphone or the like, calculates the audio synchronization feature value, and synchronizes based on the presentation time information generated in (Feature B18). In consideration of the correction information, automatic synchronization calculation with the audio synchronization feature amount of the received main channel signal is performed, and synchronization correction information (time difference information) based on the audio feature amount is calculated.

  Since the approximate synchronization position can be found from the synchronization correction information based on the presentation time information obtained by comparing the presentation time information, the amount of processing required for the automatic synchronization calculation processing using the subsequent audio synchronization feature amount can be reduced.

(Feature B20)
Based on the synchronization correction information based on the audio feature amount, the sub-receiving device performs a synchronization correction process with the main channel signal on the received sub-channel signal and reproduces it.

<Example configuration of the providing device>
Next, specific embodiments of the providing apparatus and the content reproduction system having the features B11 to B20 described above will be described.

  FIG. 22 is a diagram illustrating a configuration example of a providing apparatus that provides content having a time synchronization relationship as in the examples illustrated in (A1) to (A4) described above. In FIG. 22, parts corresponding to those in FIG. 1 are denoted by the same reference numerals, and description thereof will be omitted as appropriate.

  The providing apparatus 301 includes a reference time signal generation unit 311, a multiplexing processing unit 312, an output unit 22, a voice synchronization feature value calculation unit 23, a multiplexing processing unit 24, and an output unit 25.

  The configuration of the providing device 301 is different from the configuration of the providing device 11 in that the conversion unit 21 of the providing device 11 is replaced with a multiplexing processing unit 312 and a reference time signal generating unit 311 is newly provided. .

  The reference time signal generation unit 311 generates presentation time information indicating the content presentation timing of the main channel signal and the subchannel signal based on PCR and UTC, and supplies the presentation time information to the multiplexing processing unit 312 and the multiplexing processing unit 24. . For example, the presentation time information is PTS or the like, and this presentation time information is used on the playback side to synchronize the main channel signal and the subchannel signal.

  The multiplexing processing unit 312 converts the supplied main channel signal into a format defined by a predetermined broadcast standard or the like. Further, the multiplexing processing unit 312 generates a main transmission signal by multiplexing the main channel signal subjected to format conversion and the presentation time information supplied from the reference time signal generating unit 311, and supplies it to the output unit 22. To do. The presentation time information included in the main transmission signal is the presentation time information of the main channel signal.

  In addition, the multiplexing processing unit 24, in a time-synchronized state, the voice synchronization feature quantity supplied from the voice synchronization feature quantity calculation unit 23, the supplied subchannel signal, and the reference time signal generation unit After multiplexing the presentation time information supplied from 311, format conversion is performed as necessary to generate a sub transmission signal. The multiplexing processing unit 24 supplies the obtained sub transmission signal to the output unit 25. The presentation time information included in the sub transmission signal is the presentation time information of the subchannel signal.

  Similarly to the case of the providing apparatus 11, the multiplexing processing unit 24 may adjust the time synchronization relationship between the audio synchronization feature quantity, the subchannel signal, and the presentation time information using the main channel signal.

<Example configuration of content playback system>
Also, a content playback system that receives the main transmission signal and the sub transmission signal transmitted from the providing apparatus 301 as the main reception signal and the sub reception signal, respectively, and reproduces the main content and the sub content is configured as shown in FIG. 23, for example. Is done. In FIG. 23, parts corresponding to those in FIG. 4 are denoted by the same reference numerals, and description thereof will be omitted as appropriate.

  The content reproduction system shown in FIG. 23 includes a main reception device 341, a display unit 82, a speaker 83, a microphone 84, a sub reception device 342, a display unit 86, and a speaker 87.

  The main reception device 341 receives the main reception signal transmitted from the providing apparatus 301 and controls the reproduction of the main content obtained from the main reception signal.

  The main receiving device 341 includes an input unit 111, a separation processing unit 351, a presentation unit 352, and a reproduction processing unit 112. The configuration of the main reception device 341 is different from the configuration of the main reception device 81 in that a separation processing unit 351 and a presentation unit 352 are newly provided.

  The separation processing unit 351 separates the main reception signal supplied from the input unit 111 into a main channel signal and presentation time information of the main channel signal, supplies the main channel signal to the reproduction processing unit 112, and presents it. The time information is supplied to the presentation unit 352.

  The presentation unit 352 presents the presentation time information supplied from the separation processing unit 351 to the sub reception device 342 via a wired communication network such as the Internet or a wireless communication network. That is, the presentation time information is transmitted in response to a request from the communication partner.

  Also, the sub reception device 342 receives the sub transmission signal transmitted from the providing apparatus 301 as a sub reception signal, and controls the reproduction of the sub content obtained from the sub reception signal.

  The sub reception device 342 includes an acquisition unit 361, a presentation time information comparison unit 362, a voice synchronization feature amount calculation unit 121, a buffer 122, an input unit 123, a separation processing unit 124, a buffer 125, a synchronization calculation unit 126, and a reproduction processing unit. 127.

  The configuration of the sub reception device 342 is different from the configuration of the sub reception device 85 in that an acquisition unit 361 and a presentation time information comparison unit 362 are newly provided.

  The acquisition unit 361 acquires the presentation time information presented by the presentation unit 352 through a wired or wireless communication network using an API or the like, and supplies the presentation time information to the presentation time information comparison unit 362. That is, the acquisition unit 361 receives the presentation time information transmitted from the presentation unit 352.

  The separation processing unit 124 separates the sub reception signal supplied from the input unit 123 into the audio synchronization feature amount, the subchannel signal, and the presentation time information, and supplies the presentation time information to the presentation time information comparison unit 362. The audio synchronization feature amount and the subchannel signal are supplied to the buffer 125.

  The presentation time information comparison unit 362 compares the presentation time information supplied from the separation processing unit 124 with the presentation time information supplied from the acquisition unit 361, and synchronizes the main channel signal and the sub-channel signal. Synchronization correction information based on the presentation time information is generated and supplied to the synchronization calculator 126.

  The synchronization correction information based on the presentation time information itself can correct and synchronize the deviation between the main channel signal and the sub channel signal. However, in this example, in order to synchronize these signals with higher accuracy, the synchronization correction information based on the presentation time information is used by the synchronization calculation unit 126 to define the range of the audio synchronization feature amount read from the buffer 125. It is done. In other words, it is used for roughly synchronizing the audio synchronization feature quantity recorded in the buffer 125 and the audio synchronization feature quantity recorded in the buffer 122. As described above, by using the synchronization correction information based on the presentation time information, it is possible to perform the voice synchronization feature amount matching processing with a smaller amount of processing.

<Description of transmission processing>
Next, specific operations of the providing apparatus 301 and the content reproduction system described above will be described.

  First, the transmission processing performed by the providing apparatus 301 will be described with reference to the flowchart of FIG.

  In step S <b> 191, the reference time signal generation unit 311 generates presentation time information of the main channel signal and the subchannel signal, and supplies the presentation time information to the multiplexing processing unit 312 and the multiplexing processing unit 24.

  In step S192, the voice synchronization feature quantity calculation unit 23 performs voice synchronization feature quantity calculation processing, calculates voice synchronization feature quantities from the voice signals constituting the supplied main channel signal, and performs multiplexing processing. To the unit 24. The voice synchronization feature value calculation process performed in step S192 is the same as the voice synchronization feature value calculation process described with reference to FIG.

  In step S 193, the multiplexing processing unit 312 generates a main transmission signal by multiplexing the supplied main channel signal and the presentation time information supplied from the reference time signal generating unit 311, and outputs it to the output unit 22. Supply. At this time, the multiplexing processing unit 312 performs format conversion of the main channel signal as necessary.

  In step S194, the output unit 22 transmits the main transmission signal supplied from the multiplexing processing unit 312.

  In step S 195, the multiplexing processing unit 24 multiplexes the audio synchronization feature quantity, the subchannel signal, and the presentation time information to generate a sub transmission signal, and supplies the sub transmission signal to the output unit 25.

  That is, the multiplexing processing unit 24 multiplexes the voice synchronization feature value from the voice synchronization feature value calculation unit 23, the supplied subchannel signal, and the presentation time information supplied from the reference time signal generation unit 311. Sub-transmission signal.

  In step S196, the output unit 25 transmits the sub transmission signal supplied from the multiplexing processing unit 24, and the transmission process ends.

  As described above, the providing apparatus 301 generates presentation time information that is used in common by the main channel signal and the sub channel signal, and generates a main transmission signal and a sub transmission signal that include the presentation time information.

  As a result, on the content reproduction side, the main content and the sub-content can be synchronized with a smaller amount of processing using the presentation time information.

<Description of main content playback processing>
When the main transmission signal is transmitted from the providing device 301, the content reproduction system acquires the main transmission signal as the main reception signal and reproduces the main content. Hereinafter, the main content reproduction process by the content reproduction system will be described with reference to the flowchart of FIG.

  In step S221, the input unit 111 acquires the main reception signal and supplies it to the separation processing unit 351. For example, the input unit 111 acquires the main reception signal by receiving the main reception signal transmitted from the providing apparatus 301.

  In step S222, the separation processing unit 351 separates the main reception signal supplied from the input unit 111 into a main channel signal and presentation time information. The separation processing unit 351 supplies the separated main channel signal to the reproduction processing unit 112 and supplies the presentation time information to the presentation unit 352.

  In step S223, the reproduction processing unit 112 reproduces the main content based on the main channel signal supplied from the separation processing unit 351. In step S223, the same processing as that in step S72 in FIG. 15 is performed.

  In step S224, the presentation unit 352 presents the presentation time information supplied from the separation processing unit 351, and the main content reproduction process ends. For example, the presentation time information is transmitted to the sub reception device 342 by wireless or the like in a state synchronized with the reproduction of the main content.

  As described above, the content reproduction system obtains the main reception signal, reproduces the main content, and presents the presentation time information of the main content, that is, the main channel signal.

  In this way, by presenting the presentation time information of the main content along with the reproduction of the main content, the sub-receiving device 342 that acquires the presentation time information can perform synchronization calculation using the audio synchronization feature amount with a smaller processing amount. Will be able to do.

<Description of sub-content playback processing>
Further, in synchronization with the reproduction of the main content, the content reproduction system acquires the sub received signal and reproduces the sub content. Hereinafter, with reference to the flowchart of FIG. 26, sub-content reproduction processing by the content reproduction system will be described.

  Note that the processing in step S251 is the same as the processing in step S101 in FIG.

  In step S252, the separation processing unit 124 separates the sub reception signal supplied from the input unit 123 into a subchannel signal, a voice synchronization feature amount, and presentation time information. The separation processing unit 124 supplies the subchannel signal and the audio synchronization feature amount to the buffer 125 for recording, and supplies the presentation time information of the subchannel signal to the presentation time information comparison unit 362.

  In step S <b> 253, the acquisition unit 361 receives the presentation time information transmitted from the presentation unit 352, acquires the presentation time information of the main channel signal, and supplies the presentation time information comparison unit 362.

  In step S254, the presentation time information comparison unit 362 compares the presentation time information supplied from the separation processing unit 124 with the presentation time information supplied from the acquisition unit 361, and generates synchronization correction information based on the presentation time information. And supplied to the synchronization calculator 126.

  For example, the synchronization correction information based on the presentation time information is subjected to synchronization calculation in the synchronization calculation unit 126 among the sequence of audio synchronization feature values at each time recorded in time series in the buffer 125. The information indicates a range (hereinafter also referred to as a search range).

  This search range is the audio synchronization feature associated with the presentation time information of the sub-channel signal indicating the same time as the main content being reproduced at the present time, that is, the latest presentation time information acquired by the acquisition unit 361. A feature amount sequence for voice synchronization having a predetermined length including the amount.

  The positions of the main channel signal and the sub-channel signal having the same presentation time are signal positions synchronized with each other, that is, reproduction positions (frame positions) to be reproduced simultaneously. Therefore, by comparing the presentation time information and detecting the position of the sub-channel signal having the same presentation time information as the main channel signal, the reproduction position of the sub-content roughly synchronized with the main content being reproduced is specified. be able to.

  When the presentation time information is compared and the synchronization correction information based on the presentation time information is generated, the processes of step S255 and step S256 are performed. These processes are the same as the processes of step S103 and step S104 of FIG. Since it is the same, the description is omitted. In these processes, the audio of the main content is collected, and the audio synchronization feature amount is calculated from the audio.

  In step S257, the synchronization calculation unit 126 performs synchronization correction information generation processing, generates synchronization correction information based on the audio feature amount, and supplies the synchronization correction information to the reproduction processing unit 127. Although details of the synchronization correction information generation process will be described later, in this process, the synchronization correction information based on the presentation time information is used, and the audio synchronization feature quantity recorded in the buffer 122 and the buffer 125 are recorded. And the synchronization correction information based on the voice feature amount is generated.

  In step S258, the reproduction processing unit 127 corrects the reproduction timing of the subchannel signal recorded in the buffer 125 based on the synchronization correction information based on the audio feature amount supplied from the synchronization calculation unit 126, and performs the corrected correction. The sub contents are reproduced based on the sub channel signal. In step S258, processing similar to that in step S106 in FIG. 16 is performed.

  In this way, when the sub-content is played back in synchronization with the main content, the sub-content playback processing ends.

  As described above, the content reproduction system obtains the presentation time information of the main content and compares it with the presentation time information of the sub content included in the sub received signal, so that the synchronization correction information based on the presentation time information is obtained. Generate. Then, the content reproduction system performs matching processing on the audio synchronization feature amount included in the search range indicated by the synchronization correction information based on the presentation time information, and calculates synchronization correction information based on the audio feature amount.

  Thereby, even when the transmission paths of the main content and the sub-content are different, the synchronization correction information can be calculated with a smaller processing amount, and the contents can be reproduced in synchronization.

  That is, in the content reproduction system, the sub-receiving device 342 first synchronizes the main channel signal and the sub-channel signal roughly according to the presentation time information, and further uses the audio signal obtained by collecting the audio of the main content for audio synchronization. Calculate features.

  Then, the sub-receiving device 342 performs high-accuracy at the viewing position of the user who views the sub-contents reproduced by the sub-receiving device 342 by performing automatic synchronization calculation with the audio synchronization feature amount of the received main channel signal. Content synchronization is possible. Actually, the presentation time information is used and the range of the approximate synchronization position is narrowed down, so that the processing amount required for the automatic synchronization calculation processing by the feature amount for voice synchronization can be reduced.

  For example, in the content reproduction system, it is assumed that the display unit 82 and the speaker 83, the display unit 86 and the speaker 87 are arranged at positions away from each other, and the user is viewing the content near the display unit 86 and the speaker 87. . In such a case, it takes some time for the sound output from the speaker 83 to reach the viewing position of the user.

  Therefore, in such a case, it is difficult to synchronize the reproduction of the main content and the sub content with high accuracy at the viewing position of the user only by comparing the presentation time information. That is, for example, the audio of the main content and the audio of the sub-content are reproduced by the speaker 83 and the speaker 87 at approximately the same time, so it takes time until the audio of the main content reaches the user. Then, the user can hear the sound of the main content and the sound of the sub-content shifted.

  On the other hand, in the content reproduction system to which the present technology is applied, the sound of the main content is collected by the microphone 84 connected to the sub receiving device 342 and disposed near the sub receiving device 342, and the synchronization calculation is performed. Therefore, in the content reproduction system, it is possible to reproduce the main content and the sub content while being synchronized at the viewing position of the user. In addition, the content reproduction system compares the presentation time information to generate synchronization correction information based on the presentation time information, and limits the search range of the matching process, so that the contents can be synchronized with a smaller processing amount.

<Description of synchronization correction information generation processing>
Furthermore, the synchronization correction information generation process corresponding to the process of step S257 of FIG. 26 will be described with reference to the flowchart of FIG.

  In step S281, the frame rate conversion unit 181 and the frame rate conversion unit 182 perform frame rate conversion processing as necessary.

  That is, the frame rate conversion unit 181 reads the time series data of the audio synchronization feature value for each time interval of the main content from the buffer 122, performs frame rate conversion, that is, down-samples the audio synchronization feature value as necessary, This is supplied to the block integration unit 183.

  Also, the frame rate conversion unit 182 searches for the synchronization indicated by the synchronization correction information based on the presentation time information supplied from the presentation time information comparison unit 362 among the time-series data of the audio synchronization feature values recorded in the buffer 125. Read only the time-series data included in the range.

  Then, the frame rate conversion unit 182 performs frame rate conversion, that is, down-sample or up-sample the read audio synchronization feature quantity as necessary, and supplies the result to the block integration unit 184.

  In step S282, the block integration unit 183 and the block integration unit 184 integrate the time series data of the audio synchronization feature amount.

  Specifically, the block integration unit 183 receives the time-series data of the feature amount for audio synchronization for each time interval of the main content from the frame rate conversion unit 181. Then, the block integration unit 183 integrates the time series data of the feature quantities for voice synchronization for each supplied time section in units of blocks, with a plurality of continuous (for example, 64) time sections as one block. It supplies to the calculation part 185.

  Further, the block integration unit 184 receives supply of time-series data of feature amounts for audio synchronization for each time interval of the main content from the frame rate conversion unit 182. Then, the block integration unit 184 integrates the time series data of the feature amount for voice synchronization for each supplied time interval in units of blocks, with a plurality of continuous (for example, 64) time intervals as one block. It supplies to the calculation part 185.

  For example, as indicated by an arrow A11 in FIG. 28, in the process of step S162 in FIG. 18, each of the n blocks X (i) recorded in the buffer 122 and each of the m blocks recorded in the buffer 125. Block Y (j) was the processing target. In other words, there are n × m combinations of speech synchronization feature value blocks to be searched. More specifically, the frame rate conversion is appropriately performed on the audio synchronization feature value. However, in order to simplify the description in FIG. 28, the description will be continued assuming that the frame rate conversion is not performed.

  Here, the block Y (j) to be subjected to the matching processing is all the blocks recorded in the buffer 125 or a sufficiently wide range of blocks.

  In FIG. 28, i is the index of the block of the feature value for voice synchronization obtained by the feature amount calculation unit 121 for voice synchronization, and j is the feature value for voice synchronization included in the sub-received signal. The block index.

  X (i) represents time-sequential data of the voice synchronization feature quantity of the block of index i among the voice synchronization feature quantities obtained by the voice synchronization feature quantity calculation unit 121, and Y (j) represents , Represents the time-sequential data of the voice synchronization feature quantity of the index j block among the voice synchronization feature quantities included in the sub received signal.

  On the other hand, in step S282, m ′ blocks included in the search range indicated by the synchronization correction information based on the presentation time information among the m blocks recorded in the buffer 125, as indicated by an arrow A12. Only blocks are subject to matching processing. That is, it is a target of similarity calculation in the similarity calculation unit 185.

In this example, PTS _i represents presentation time information, and the position indicated by the presentation time information is the position of the main content being played back at the present time. A range of a predetermined length including a position corresponding to the presentation time information of the sub-content at the same time as the presentation time information of the main content, that is, a range including m ′ blocks is set as the search range. Accordingly, there are n × m ′ combinations of blocks of the feature amount for voice synchronization to be searched.

  As described above, if the synchronization correction information based on the presentation time information obtained by comparing the presentation time information is used, it is possible to limit the range of the audio synchronization feature amount to be subjected to the matching process to the minimum necessary. The processing time required for searching for similarity calculation can be greatly reduced.

  Returning to the description of the flowchart of FIG. 27, when the time-series data of the feature amount for audio synchronization is integrated, the process proceeds to step S283. Then, the processing of step S283 and step S284 is performed and the synchronization correction information generation processing ends. Since these processing are the same as the processing of step S163 and step S164 of FIG. 18, the description thereof is omitted. When the synchronization correction information generation process ends, the process proceeds to step S258 in FIG.

  As described above, the content reproduction system generates the synchronization correction information based on the audio feature amount by using the audio synchronization feature amount in the search range indicated by the synchronization correction information based on the presentation time information. As a result, the synchronization correction information can be generated robustly with a smaller amount of processing.

  Further, the providing apparatus 301 shown in FIG. 22 and the content reproduction system shown in FIG. 23 are also applicable to the systems shown in FIGS.

  By the way, the above-described series of processing can be executed by hardware or can be executed by software. When a series of processing is executed by software, a program constituting the software is installed in the computer. Here, the computer includes, for example, a general-purpose computer capable of executing various functions by installing a computer incorporated in dedicated hardware and various programs.

  FIG. 29 is a block diagram illustrating a hardware configuration example of a computer that executes the above-described series of processing by a program.

  In a computer, a CPU (Central Processing Unit) 501, a ROM (Read Only Memory) 502, and a RAM (Random Access Memory) 503 are connected to each other by a bus 504.

  An input / output interface 505 is further connected to the bus 504. An input unit 506, an output unit 507, a recording unit 508, a communication unit 509, and a drive 510 are connected to the input / output interface 505.

  The input unit 506 includes a keyboard, a mouse, a microphone, an image sensor, and the like. The output unit 507 includes a display, a speaker, and the like. The recording unit 508 includes a hard disk, a nonvolatile memory, and the like. The communication unit 509 includes a network interface or the like. The drive 510 drives a removable medium 511 such as a magnetic disk, an optical disk, a magneto-optical disk, or a semiconductor memory.

  In the computer configured as described above, the CPU 501 loads the program recorded in the recording unit 508 to the RAM 503 via the input / output interface 505 and the bus 504 and executes the program, for example. Is performed.

  A program executed by the computer (CPU 501) can be provided by being recorded on a removable medium 511 as a package medium, for example. The program can be provided via a wired or wireless transmission medium such as a local area network, the Internet, or digital satellite broadcasting.

  In the computer, the program can be installed in the recording unit 508 via the input / output interface 505 by attaching the removable medium 511 to the drive 510. Further, the program can be received by the communication unit 509 via a wired or wireless transmission medium and installed in the recording unit 508. In addition, the program can be installed in advance in the ROM 502 or the recording unit 508.

  The program executed by the computer may be a program that is processed in time series in the order described in this specification, or in parallel or at a necessary timing such as when a call is made. It may be a program for processing.

  The embodiments of the present technology are not limited to the above-described embodiments, and various modifications can be made without departing from the gist of the present technology.

  For example, the present technology can take a configuration of cloud computing in which one function is shared by a plurality of devices via a network and is jointly processed.

  In addition, each step described in the above flowchart can be executed by being shared by a plurality of apparatuses in addition to being executed by one apparatus.

  Further, when a plurality of processes are included in one step, the plurality of processes included in the one step can be executed by being shared by a plurality of apparatuses in addition to being executed by one apparatus.

  Moreover, the effect described in this specification is an illustration to the last, and is not limited, There may exist another effect.

  Furthermore, this technique can also be set as the following structures.

(1)
A feature amount calculation unit that extracts a feature amount from the audio signal of the first content;
By comparing the feature amount acquired in a synchronized state with the second content having a time synchronization relationship with the first content and the feature amount extracted by the feature amount calculation unit An information processing apparatus comprising: a synchronization calculation unit that generates synchronization correction information based on an audio feature amount for reproducing the second content in synchronization with the first content.
(2)
The information processing apparatus according to (1), wherein the feature amount calculation unit extracts the feature amount from the audio signal obtained by collecting the reproduced audio of the first content.
(3)
(1) or (1) further comprising a first input unit that acquires the second content and the feature amount associated with the second content in a state in which the second content is synchronized. The information processing apparatus according to 2).
(4)
The information processing apparatus according to (3), wherein the second content and the feature amount are transmitted to the information processing apparatus at a timing that takes into account a difference in arrival time with respect to the first content.
(5)
The first input unit requests transmission of the second content and the feature amount, and receives the second content and the feature amount transmitted in response to the request. Information processing device.
(6)
The information processing apparatus according to (5), further comprising: a second input unit that requests transmission of the first content and receives the first content transmitted in response to the request.
(7)
The feature quantity calculation unit extracts the feature quantity from the audio signal for one reproduced first content,
The synchronization calculation unit compares each of the feature amounts of the plurality of first contents associated with the second content with the feature amount extracted by the feature amount calculation unit. , Specifying the reproduced first content, and generating synchronization correction information based on the audio feature amount for reproducing the identified first content and the second content in synchronization (2) Thru | or the information processing apparatus as described in any one of (6).
(8)
The information processing apparatus according to any one of (2) to (7), further including a reproduction processing unit that controls reproduction of the second content.
(9)
The information processing apparatus according to (8), wherein the reproduction processing unit corrects the reproduction position of the second content based on synchronization correction information based on the audio feature amount.
(10)
An acquisition unit for acquiring presentation time information of the first content;
A comparison unit that compares the presentation time information of the first content with the presentation time information of the second content and generates synchronization correction information based on the presentation time information;
The synchronization calculation unit includes the feature amount included in a range indicated by the synchronization correction information based on the presentation time information in the acquired feature amount series, and the feature amount extracted by the feature amount calculation unit. The information processing apparatus according to any one of (1) to (9), wherein synchronization correction information based on the audio feature amount is generated.
(11)
The synchronization calculation unit uses the acquired feature value or the feature value calculation unit so that a frame rate of the acquired feature value matches the feature value extracted by the feature value calculation unit. The information processing apparatus according to any one of (1) to (10), wherein frame rate conversion is performed on at least one of the extracted feature values, and then the feature values are compared.
(12)
A feature amount calculating step for extracting a feature amount from the audio signal of the first content;
The feature amount acquired in synchronization with the second content having a time synchronization relationship with the first content is compared with the feature amount extracted by the processing of the feature amount calculation step. Thus, an information processing method including a synchronization calculation step of generating synchronization correction information based on an audio feature amount for reproducing the second content in synchronization with the first content.
(13)
A feature amount calculating step for extracting a feature amount from the audio signal of the first content;
The feature amount acquired in synchronization with the second content having a time synchronization relationship with the first content is compared with the feature amount extracted by the processing of the feature amount calculation step. Thus, a program for causing a computer to execute a process including: a synchronization calculation step of generating synchronization correction information based on an audio feature amount for reproducing the second content in synchronization with the first content.
(14)
A feature amount calculation unit that extracts a feature amount from the audio signal of the first content;
The second content having a time synchronization relationship with the first content, and the feature amount associated with the second content in a state of being synchronized with the second content An information processing apparatus comprising: a first output unit.
(15)
The information processing apparatus according to (14), further including a second output unit that outputs the first content.
(16)
The information processing apparatus according to (15), wherein the first output unit outputs the second content and the feature amount at a timing in which a difference in arrival time from the first content is considered.
(17)
The information output according to (15), wherein when the transmission of the second content and the feature amount is requested, the first output unit outputs the second content and the feature amount in response to the request. apparatus.
(18)
The information processing apparatus according to (17), wherein when the transmission of the first content is requested, the second output unit outputs the first content in response to the request.
(19)
The feature amount calculation unit extracts the feature amount from the audio signal for a plurality of the first contents,
The information processing apparatus according to any one of (14) to (18), wherein the first output unit outputs the feature quantities of the plurality of first contents in association with the second contents.
(20)
The feature amount calculation unit downsamples the feature amount,
The information processing apparatus according to any one of (14) to (19), wherein the first output unit outputs the second content and the downsampled feature amount.
(21)
A feature amount calculating step for extracting a feature amount from the audio signal of the first content;
The second content having a time synchronization relationship with the first content, and the feature amount associated with the second content in a state of being synchronized with the second content An information processing method comprising: an output step.
(22)
A feature amount calculating step for extracting a feature amount from the audio signal of the first content;
The second content having a time synchronization relationship with the first content, and the feature amount associated with the second content in a state of being synchronized with the second content A program for causing a computer to execute a process including an output step.

  DESCRIPTION OF SYMBOLS 11 Provision apparatus, 22 Output part, 23 Voice synchronization feature-value calculation part, 24 Multiplexing process part, 25 Output part, 81 Main receiving apparatus, 85 Sub receiving apparatus, 111 Input part, 112 Playback processing part, 121 For voice synchronization Feature amount calculation unit, 123 input unit, 126 synchronization calculation unit, 127 reproduction processing unit, 311 reference time signal generation unit, 352 presentation unit, 361 acquisition unit, 362 presentation time information comparison unit

Claims (5)

A band dividing unit for dividing the sound signal included in the first content;
A periodicity detection unit that detects, for each band, periodicity information of the acoustic signal that is band-divided by the band-division unit;
A periodicity information integration unit that integrates the periodicity information for each band detected by the periodicity detection unit for all bands;
Detecting a peak position of periodicity information integrated by the periodicity information integration unit, and generating peak information;
A down-sampling unit that uses the peak information of a plurality of time intervals generated by the peak detection unit as information of one time interval;
A signal processing apparatus comprising: an output unit that outputs information down-sampled by the down-sampling unit as a synchronization feature amount when synchronizing the first content and the second content to be synchronized. The signal processing apparatus according to claim 1, wherein the acoustic signal is an acoustic signal obtained by collecting sound of the reproduced first content. A playback processing unit for controlling playback of the second content;
The reproduction processing unit is configured to compare the second content based on the synchronization correction information obtained by comparing the feature amount of the first content associated with the second content with the feature amount for synchronization. The signal processing apparatus according to claim 1, wherein the reproduction position of the content is corrected. The signal processor
A band dividing process for dividing the sound signal included in the first content;
Periodicity detection processing for detecting, for each band, periodicity information of the acoustic signal that has been band-divided by the band-division processing;
Periodicity information integration processing for integrating the periodicity information for each band detected by the periodicity detection processing for all bands;
A peak detection process for detecting peak positions of the periodicity information integrated in the periodicity information integration process and generating peak information;
Down-sampling processing in which the peak information of a plurality of time intervals generated by the peak detection processing is information of one time interval;
A signal processing method for executing an output process for outputting information down-sampled by the down-sampling process as a synchronization feature amount when synchronizing the first content and the second content to be synchronized. A band dividing process for dividing the sound signal included in the first content;
Periodicity detection processing for detecting, for each band, periodicity information of the acoustic signal that has been band-divided by the band-division processing;
Periodicity information integration processing for integrating the periodicity information for each band detected by the periodicity detection processing for all bands;
A peak detection process for detecting peak positions of the periodicity information integrated in the periodicity information integration process and generating peak information;
Down-sampling processing in which the peak information of a plurality of time intervals generated by the peak detection processing is information of one time interval;
A program that causes a computer to execute a process including: an output process that outputs information down-sampled by the down-sampling process as a synchronization feature amount when synchronizing the first content and the second content to be synchronized.

Images