JP2003309814A - Moving picture reproducing apparatus, moving picture reproducing method, and its computer program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a moving picture reproducing apparatus for accurately detecting a voice utterance period in a voice signal uttered by a person and remarkably reducing a time required for browsing by a user while maintaining a synchronization relation between video and voice with fidelity according to the detected voice period. <P>SOLUTION: The moving picture reproducing apparatus discriminates a period A for denoting a voice utterance period of a person and periods B other than the period A on the basis of the voice signal included in moving picture data, reproduces a moving picture together with reproduced voice at a high speed being a non-multiple speed or a prescribed speed (e.g. 1.5 time or double speed of the non-multiple reproduction speed) at which a user can easily grasp the contents of the data for the period A and performs high speed moving picture reproduction together with reproduction voice in a small voice amount or in silence at a speed higher than the prescribed speed (e.g. 5 to 10-speed of the non-multiple speed) for the periods B on the basis of the moving picture data. In this case, the moving picture reproduction speed can be adjusted depending on attribute information of the user registered in a user profile 14. <P>COPYRIGHT: (C)2004,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to the field of moving image reproduction technology involving audio reproduction.

[0002]

2. Description of the Related Art Conventionally, in a moving picture reproducing apparatus which reproduces sound, such as a video tape recorder, a user views the whole moving picture (that is, the whole content to be reproduced) in a short time when executing reproduction. To enable that,
It has a double speed playback function and a fast forward function.

Further, in a video tape recorder, which is a typical moving picture reproducing apparatus, in recent years, when a double speed reproduction of a recording medium is executed, a first voice section in which the energy of sound is equal to or higher than a predetermined threshold value and the predetermined track. By detecting a second voice section less than the threshold value and continuously reproducing while performing pitch conversion of the voice signal in the first voice section, the reproduced voice is eroded while eroding the second voice section. Although it is a bit quicker for the user, there is also proposed a technology capable of reproducing a storage medium at a double speed while accompanied by a reproduced sound capable of understanding the contents.

[0004]

However, if the partial pitch conversion processing of the audio signal is performed as described above, the synchronous relationship between the audio and the video cannot always be maintained during the reproduction of the moving picture (fast-playing of the moving picture). For example, since the video in which the person speaking in the reproduced video is not synchronized with the reproduced audio, the reproduction may be unnatural to the human sense, and the user may feel uncomfortable.

Further, for example, in Japanese Unexamined Patent Publication No. 10-32776 and Japanese Unexamined Patent Publication No. 9-243351, a silent state is detected on the basis of voice energy, and a sound section other than the detected silent state is produced by a person. By considering
Techniques for summarizing videos have also been proposed.
However, in a moving image in which human voice is dominant throughout the program, such as a news program, background noise can be detected to some extent, although voice segment of human voice can be detected based on voice energy. This method is not practical in an environment with background music and background music.

Further, in the prior arts prior to the above patent publication, there have been proposed many techniques for detecting a voice and reproducing a moving image in consideration of the detected voice, most of which are thresholding the energy of the sound. The voice is detected by doing. There is a problem due to the ambiguity of Japanese in this background, because "human voice" is also called "voice", and general sounds including human voice are also called "voice".
It is inappropriate to collectively refer to such conventional sound energy thresholding as a true "voice detection".

Further, in Japanese Unexamined Patent Publication No. 9-247617, "characteristic points such as voice information" are detected by calculating a FFT (Fast Fourier Transform) spectrum of a voice signal to find a singular point, and its volume. The technique of analyzing is proposed. However, in the method using the FFT spectrum, when the audio signal to be reproduced includes so-called background music having a wide band spectrum distribution, it is necessary to detect a human voice from the background music. It is difficult.

As described above, in the conventional moving image reproduction involving voice, there is a problem that the detection of the voice section is convenient and inaccurate as described above, and further, the summary of the moving image using the detection result is generated. When creating or playing back at double speed,
During reproduction, there is a problem that the synchronization relationship between video and audio cannot be maintained.

Generally, it is not easy for a user such as an old man or a child to use various devices,
Moreover, it is known that it is difficult to understand the content of a voice that is emitted at a high speed. Therefore, for such a user, the optimum conditions for the reproduction are different from those of a general user when performing a quick preview (shortening reproduction) of the contents such as the double speed reproduction in the moving picture reproducing apparatus such as the above-mentioned tape recorder.

Further, even for a user who has a weak visual acuity, a user who has a weak hearing ability for a fast voice, or a foreign user who does not speak the voice to be played back in his or her native language, the above-mentioned moving picture playback apparatus can be used to reproduce the contents such as double speed. When performing quick viewing (shortened playback), the optimum conditions for playback differ from those of general users.

Therefore, the present invention accurately detects the voice section uttered by a person, and faithfully maintains the synchronization relationship between the image and the voice in accordance with the detected voice section, while significantly shortening the time required for the user to browse. An object is to provide a moving image reproducing device, a moving image reproducing method, and a computer program thereof.

[0012]

In order to achieve the above object, a moving picture reproducing apparatus according to the present invention is characterized by the following configuration.

That is, a moving picture reproducing apparatus capable of reproducing moving picture information including a sound signal at a high speed, wherein a first sound section representing a human utterance period based on the sound signal contained in the moving picture information, and Other than the second voice section, and based on the moving picture information, the first voice section performs high-speed moving picture reproduction accompanied by reproduced sound at a predetermined speed at which the user can grasp the contents. On the other hand, the second voice section is provided with a quick-view playback unit that performs at least high-speed moving image playback at a speed higher than the predetermined speed.

[0014] In a preferred embodiment, the quick-view reproduction means may perform a moving image reproduction at a speed higher than the predetermined speed and with a reproduced sound of at least a small volume in the second audio section.

Alternatively, in another preferred embodiment, the quick-view reproduction means may reproduce the moving image without voice at a speed higher than the predetermined speed in the second voice section.

In any of the above device configurations, the voice section determination means extracts a voice pitch corresponding to a vocal cord vibration based on the voice signal, and determines the first voice section based on the extracted voice pitch. Good to judge.

Further, in any of the above device configurations, the voice section determination means may have a vocal cord frequency range that can exist from a signal obtained by filtering the voice band emitted by a person included in the voice signal. It is characterized in that the first vowel section is determined by detecting the dominant vowel part of the human voice by extracting the pitch of the above and integrating the detected vowel parts.

In addition, for example, the voice section determination means, when determining the first voice section based on the voice signal, corrects means for integrally correcting a plurality of first voice sections that are close to each other on a time axis. In this case, the correction unit detects a scene change point included in the moving image information, and detects the scene change point included in the moving image information, and detects the scene change point from the start point of the first voice section of interest. If the time interval (that is, the distance on the time axis) between the near scene change point located closest in time and the start point thereof (that is, the distance on the time axis) is equal to or less than a predetermined threshold value, the first The start point of the voice section may be corrected by replacing it with information corresponding to the nearby scene change point.

Further, for example, the quick-view reproduction means calculates a time required for the high-speed moving image reproduction based on the length of the first voice section, the reproduction speed of the section, and the length of the second voice section. Along with this, the calculated required time is presented to the user, and in this case, the quick-view reproduction means is adapted to present the required time, and in response to the presentation of the required time.
And when the operation of changing the reproduction speed of the second voice section is performed by the user, based on the changed reproduction speed,
It is preferable to include adjustment means for adjusting the required time.

In a preferred embodiment, attribute information (eg, age, language used, dynamic visual acuity, etc.) about each user is targeted for users who can use the moving image reproducing apparatus.
And a user profile in which the speed of hearing of fast voice, etc.) is registered,
It is preferable that the reproduction speeds of the first and second voice sections are automatically determined according to the attribute information regarding the specific user, which is registered in the profile.

The same object can also be achieved by a moving picture reproducing method corresponding to the moving picture reproducing apparatus having the above-mentioned constitutions.

The same object can also be achieved by a program code for realizing a moving picture reproducing apparatus and method having the above-mentioned configurations by a computer, and a computer-readable storage medium storing the program code. .

[0023]

BEST MODE FOR CARRYING OUT THE INVENTION An embodiment of a moving image reproducing apparatus according to the present invention will be described in detail below with reference to the drawings.

First, an outline of the operation of the moving picture reproducing apparatus according to this embodiment will be described with reference to FIG.

FIG. 1 is a diagram showing a conceptual diagram of a moving picture fast-viewing algorithm in the moving picture reproducing apparatus according to this embodiment.

As shown in FIG. 1, the moving picture reproducing apparatus according to the present embodiment is roughly classified into a moving picture fast-view index creating section 1
00 and the moving image quick playback unit 200.

<Motion picture fast-view index creation unit 100> In the motion picture quick-view index creation unit 100, the motion picture data read from the motion picture data storage unit 10 is converted into video data and audio data (audio signal) in the video / audio separation process (step S101). To the audio signal,
A voice segment estimation process (step S102) and a voice segment correction process (step S103) are performed, and a video change degree calculation process (step S105) and a scene change point detection process (step S106) are performed on the video data. The quick view reproduction section information is generated by the quick view reproduction section correction processing (step S104), and the generated information is stored in the moving picture quick view index storage unit 11.

That is, the voice section estimation process (step S10)
In 2), the low-pass filter is applied to the audio signal obtained in the video / audio separation process (step S101) to obtain the zero-crossing point of the audio signal, and the zero-crossing point is set as the start point. And a terminal point are formed, and when the signal energy of the adjacent small segment is small, the small segment is combined with the immediately preceding small segment to obtain 1
One small segment is determined. Here, the zero crossing point is a point where the waveform of the filtered audio signal crosses the zero level which is the reference signal level.

Attribute information representing the property is attached as a label to each small segment thus determined. This label always contains a label for the voice pitch, and within the voice pitch segment,
It also has voice pitch cycle information (details will be described later).

In the present embodiment, upon speech detection, the distance between segments having adjacent speech pitch labels (that is, the time interval, the same applies hereinafter) is thresholded on the basis of the segment group having such speech pitch labels. In addition, by using the continuity of the voice pitch period information individually included in the voice pitch segments, and further considering both the continuity of the voice pitch period and the distance between the segments, a discrete voice is obtained. A voice segment is accurately detected by integrating a plurality of small segments having pitch labels as one segment.

Further, a voice section correction process (step S10)
In 3), based on the processing result (voice section detected from the voice signal) in the voice section estimation process (step S102), a plurality of adjacent voices are stored so that the person (user) does not feel uncomfortable during voice reproduction. Corrected voice section information by correcting a voice section (hereinafter, referred to as “human voice section” or section A) representing a human vocalization period, which is newly reproduced by integrating the sections. To get.

For example, as a bad aspect in high-speed moving image reproduction, when the interval between two adjacent sections A is small, when reproducing the moving image, those audio sections are at a speed at which a person can hear and grasp the contents. In addition to performing double speed reproduction accompanied by sound (for example, double speed reproduction), for a section other than a person's voice section (hereinafter referred to as section B), the person grasps the content by watching the reproduced image when reproducing the moving image. When you play back at a high speed with a high magnification that allows you to
It is difficult for general users to hear.

Therefore, in this embodiment, in the voice section correction process (step S103), the intervals of the voice sections of the person are taken into consideration, and if the intervals satisfy a predetermined condition, the voice section groups of a plurality of persons are selected. By integrating, the above-mentioned listening difficulty is eliminated. Here, as the predetermined condition, for example, it is easiest to set that the interval of the human voice section is equal to or smaller than a predetermined threshold value.

Further, a video change degree calculation process (step S10)
In 5), with respect to the video data obtained in the video / audio separation process (step S101), Japanese Patent Application Laid-Open No. 2000-23563
By calculating similarity between frames by performing similarity comparison processing between frames described in Japanese Patent Publication No.
Video change information is generated.

In general, when a video transition exists in moving image data including an audio signal and a voice section starts immediately after that, when the moving image is reproduced, the image of the beginning portion of the scene is reproduced only at a momentary high speed. Since the reproduced video by the double speed reproduction accompanied by the audio is performed at a speed that can be heard and grasped by a person, the user feels uncomfortable as if the video flickers.

Therefore, in the present embodiment, in the scene change point detection processing (step S106), for example, the technique of detecting the scene change point disclosed in Japanese Patent Laid-Open No. 2000-235639 by the applicant of the present application is adopted. Thus, the scene change point group (scene change point information) is detected based on the video change information obtained in the video change degree calculation process (step S105).

Then, in the quick-view reproduction section correction processing (step S104), the distance is less than or equal to a predetermined threshold in time and closest to the beginning of the voice section after the voice section correction processing in step S103. In this case, by replacing the beginning of the voice section with the information corresponding to the scene change point detected in step S103, the user's discomfort can be removed.

<Movie Fast Playback Unit 200> Next, the movie fast playback unit 200 performs a quick movie playback process (step S10).
In 7), the reproduced video is reproduced using the display 12, and the reproduced sound is reproduced using the speaker 13. When the moving image is reproduced by this moving image quick reproduction process, the time required for reproduction is displayed in step S108 based on the rapid reproduction section information read from the moving image quick index storage unit 11, and a step corresponding to the display is displayed. The feedback of the reproduction condition desired by the user set in S109 and the reproduction condition based on the user profile 14 are integrally determined to finally set the quick view reproduction condition, and based on the set quick view reproduction condition. , Video data storage unit 1
Video reproduction of the video data read from 0 is performed.

At this time, in the present embodiment, for the section A, the double speed reproduction accompanied by the voice is performed at a speed that allows the user to understand the content when the user hears the reproduced voice. In other words, high-speed double-speed reproduction is performed within a range in which the user can understand the content by watching the reproduced video.

Here, the double speed reproduction in the above section A,
That is, it is known from the result of the experiment conducted by the applicant of the present application that the reproduction at a speed at which a person can hear and understand the content should be up to double speed in the experiment, preferably about 1.5 speed. On the other hand, with respect to the section B, the reproduced image is reproduced at a high speed with a high magnification within a range where a person can see and understand the content. ,
It has been found that it is desirable to set the speed to 5 times or more.

It is known that, when the section B is reproduced at a high speed and at a high speed, a sound of "curculer" is generally produced. Therefore, in step S107, when the section B is reproduced at a high speed, such a sound is produced. If you do not want to listen, it is conceivable to mute the audio reproduction to make it silent or to reduce the volume during reproduction.

Regarding the reproduction speed of the section A, the reproduction speed of the section B, and the volume at the time of reproduction, the simplest implementation method is to decide beforehand how to process the audio in the moving image quick reproduction processing (step S107). In addition, there is a method of allowing the user to variably set the reproduction speed.

However, it is generally known that it is not easy for a user such as an old man or a child to use various devices, and it is difficult to understand the contents when a high speed voice reproduction is performed. Therefore, it is preferable to easily perform the double speed reproduction with a slightly low magnification without performing the troublesome speed adjustment. Similarly, regardless of age, users with low visual acuity (visually impaired), especially those with weak visual acuity and hearing, especially those with low-speed hearing (audible people), or reproduced voices in their native language It is known that foreign users and the like do not understand the content when high-speed voice reproduction is performed, and there is also an optimum reproduction speed for these users.

Therefore, in the present embodiment, information such as the age and language of the user, understandable language, visual acuity and hearing, and attribute information about the user such as reference reproduction conditions preferred by each user are stored in the user profile 14. Stored in advance, and referring to the profile 14 in the video fast-viewing reproduction process (step S107), the human utterance section (section A) and the section excluding the human utterance section are selected according to the target user. It is possible to determine the playback speed for each of the (section B) and perform quick movie playback in which the content can be easily understood according to the individual.

Further, as described above, even in the case where the mute of the sound or the volume of the sound is reduced during the high-speed double-speed reproduction of the section B, such setting is described in the profile 14 in advance so that the individual user It is possible to perform a comfortable quick movie playback according to the requirements.

Further, for the elderly and users with handicaps in dynamic visual acuity, the reproduction speed of the section A may be set to be slower than the normal speed, and the section B may be deviated from the viewpoint of the original fast-view reproduction. By setting the playback speed of the above to be equal to or faster than 1 × speed, the user needs a shorter overall playback speed than the slow playback speed in which the user can grasp the voice content of the section A. It is possible to browse the moving image (that is, the moving image data stored in the moving image data storage unit 10) in time.

Further, for a user who has a handicap in understanding the content of a fast voice and a user who is not fluent in the language of the voice content, although it may be out of the original viewpoint of quick playback, the playback speed of the section A is the same. In addition to setting the speed lower than the speed, the playback speed of the section B is up to 10 times, preferably 5 times or more, and while performing the low-speed playback in which the user can grasp the audio content of the section A, the entire section is played as a whole. It is possible to browse a moving image (that is, moving image data stored in the moving image data storage unit 10) in a shorter required time as compared with the case of low speed reproduction. here,
To determine whether or not the user is fluent in the language of the audio content, the identification information (preferred language in Table 4 described later) stored in the profile 14 described above is compared with the language type information of the audio included in the moving image to be reproduced. It can be done by doing things.

As a procedure for selecting the user profile 14, for example, the user profile list is displayed on the profile selection screen displayed on the display 12, and the user selects a remote control terminal (not shown) from the list.
It is conceivable that the profile is selected in accordance with the operation of, and an automatic profile selection method using a personal recognition technique such as fingerprint, voice print, face recognition or the like may be adopted.

By the way, in the case of performing the optimum quick-view reproduction for each user as described above, it is attempted to use the free time to perform the quick-view to find out how long the moving picture originally originally can be played. Information that is important to users.

Therefore, in this embodiment, step S108.
In Section 1, the playback time of the section A is calculated by dividing the total length of the section A by the playback speed, and for the section B, the playback speed of the section B is calculated by dividing the total length by the playback speed. As the time required,
The sum of these two calculated values is calculated and presented to the user together with the time required for reproducing the original moving image at the same size. Furthermore,
The reproduction time is adjusted so that it is close to the reproduction time desired by the user by designating the reproduction speed of the section A and the reproduction speed of the section B so that the reproduction time falls within the desired reproduction time by the user. It is possible to

By the way, regarding the relationship between the preset user profile 14 and the desired playback speed instructed by the user, the movie quick-view playback automatically calculated using the profile 14 in step S108 as described above. The user who has seen the time required to specify the reproduction speeds of the section A and the section B in step S109 through a predetermined man-machine interface further determines the time required to reproduce the desired movie fast (the reproduction speed. (Information) is set, the set playback speed information is newly stored in the profile in order to save it within the set time, automatically or after confirming with the user. It is possible to perform quick-view playback of a moving image that is easy to understand according to the preference of each user while reflecting the operation information of.

In addition, in the above user profile,
Further, if the volume of the sound at the time of reproduction in the section B is designated in advance, or if the user designates it through a predetermined man-machine interface, the individual volume is reflected while reflecting the designated volume information. It is possible to perform quick-view playback that is easy to understand according to the user's preference.

<Details of Operation of Video Reproducing Device> Hereinafter, details of the operation of the video reproducing device according to the present embodiment outlined above will be described. In the following description, the quick view playback section information is created as an index for performing the quick view playback for the recorded video data stored in the video data storage unit 10, and the created video information is used to create the quick view playback section information. An example will be described in which the quick-view reproduction is performed.

In the present embodiment, as described above, the post-processing after the video / audio separation processing of step S101 is roughly divided into the video quick-view index creation processing by the video quick-view index creation section 100 and the video quick-view playback section 200. There is a movie quick-view playback process.

<Motion picture quick-view index creation unit 100> FIG. 2 is a block diagram showing an algorithm for detecting a voice section (section A) representing a person's utterance period, which is performed by the motion picture quick-view index creation unit 100.
(Automatic gain control) 21, low-pass filter 2
2. A zero crossing detection unit 23a, 23b, a voice segmentation unit 24, a voice pitch detection unit 25, a voice labeling unit 26, a voice energy calculation unit 27, and a voice section estimation unit 28.

FIG. 3 is a flow chart showing the outline of the processing based on the algorithm shown in FIG. 2. The procedure for detecting the section A will be described with reference to this flow chart. First, in step S301, the audio signal is converted into a plurality of small segments. And in step S302, speech labeling is performed to express the acoustic features of these small segments. At that time, in step S303, a robust vowel candidate is detected by detecting the voice pitch, and finally, in step S303.
At 304, the human voice section (section A) is estimated based on the voice pitch detection result.

That is, video / audio separation processing (step S10
The audio signal separated from the video data by 1) is A
The voice energy is normalized by a GC (auto gain control) 21. A well-known configuration may be adopted for the AGC21 configuration.
It is possible to adopt a configuration in which normalization is performed with the sound having the maximum signal level as a reference throughout.

The normalized voice signal is filtered by the low-pass filter 22 to obtain a voice signal component in a band suitable for analysis processing performed at a later stage and an original voice signal having a band necessary for unvoiced consonant recognition. Branch to.

(Voice Segmentation) First, in the voice signal that has passed through the low-pass filter 22, after the zero crossing point is obtained by the zero crossing point detecting section 23a, the zero crossing point is used as a reference.
In the audio segmenting unit 24, the audio segmentation unit 24 tentatively divides it into small parts called "small segments". This process corresponds to step S301 in FIG.

Here, the reason why the low-pass filter 22 is used for segmentation is that the reference for small segments is unvoiced consonants,
This is because it is a unit of voiced consonant, voice pitch, and the like, and unvoiced consonants and the like are adversely affected by the influence of high frequencies.

Now, the voice segmenting unit 24 divides the voice signal into small segments with reference to the zero crossing point tentatively obtained for the voice signal, and the small segment has the following two-condition rule. 1: The start point and the end point of the small segment are zero crossings. Rule 2: When the energy of the small segment is small,
Combine with the previous small segment.

The sound energy P is given to a small segment f (x) whose starting point is X1 and whose ending point is X2,

[0063]

[Equation 1] It is defined that the following mathematical expression (1) is satisfied.

When the calculated sound energy P is less than or equal to the predetermined threshold value Eth1, the currently targeted small segment f (x) is integrated with the immediately preceding small segment. The sound energy P may be calculated by using the squared energy of f (x) instead of accumulating the absolute values of the small segments f (x) according to Expression (1).

FIG. 4 is a diagram for explaining the small segment joining process performed in this embodiment.

In FIG. 4, FIG. 4A exemplifies the audio signal level at which a plurality of zero cross points (Zero cross points) are obtained by the zero cross point detecting section 23a. Also, FIG.
In (b), the detected zero-crossings, a plurality of small segments set by applying the above-mentioned rules 1 and 2 are shown by individual vertical lines,
The two small segments indicated by the arrows indicate that they are integrated into one small segment according to the above-mentioned rule 2.

(Voice Labeling Processing) Zero Crossing Detection Unit 23b
Then, the average number of zero crossings at which the voice signal waveform in which the voice energy is normalized by the AGC 21 intersects with the reference zero level is obtained, and further the voice energy calculation unit 27 obtains the average energy. Against
The voice labeling unit 26 calculates the start point, the end point, the average number of zero crossings, and the average energy, and stores the calculated values as the feature amount of the small segment. This process corresponds to step S302 in FIG.

However, the average number of zero crossings and the average energy are calculated by the following formulas using the segment length SegLen.

(Average number of zero-crossings) = (number of zero-crossing points of original speech signal included in small segment) / SegLen, (average energy) = (low-pass filtered speech signal included in small segment) Energy) / SegLen.

Further, the small segments are classified into five types of categories, and labels showing the categories are given. In the present embodiment, the types of labels that can be given to individual small segments include silence, unvoiced consonants, voiced consonants, voice pitch, and noise.

Next, it is determined by the procedure shown in FIG. 5 which label the small segment of interest corresponds to.

FIG. 5 is a flow chart showing the process of voice labeling performed in this embodiment, and shows the procedure of the process performed by the voice labeling unit 26.

In step S501 of the figure, the average number of zero crossings AveZeroCrossRate and the average energy AveEnergy are read in as feature quantities of the small segment of interest (small segment to be processed).

In this embodiment, as the label judgment condition,
The following thresholds are provided, but these thresholds are all constants.

• Threshold representing maximum energy of silence: SileceEnergyMax, • Minimum energy threshold of unvoiced consonants: ConHEnergyLow, • Maximum energy threshold of unvoiced consonants: ConHEnergyMax, • Minimum energy of voiced consonants Threshold: ConLEnergyLow, ・ Maximum energy threshold for voiced consonants: ConLEnergyMax, ・ Minimum zero-crossing threshold for unvoiced consonants: ConHZeroCrossRateLow, ・ Maximum zero-crossing threshold for voiced consonants: ConLZeroCrossRateMax, where SileceEnergyMax ＞ ConHEnergyLow shall be satisfied.

In step S502, it is determined whether the feature amount read in step S501 satisfies a predetermined silent condition. Here, the silent label condition is: ((AveEnergy <SileceEnergyMax) AND (AveZeroCros
sRate <ConHZeroCrossRateLow)), or ((AveEnergy <ConHEnergyLow) AND (AveZeroCrossR
ate> ConHZeroCrossRateLow)), Then, in step S503, when the silent label condition is satisfied, the silent label is associated with the small segment of interest and stored.

On the other hand, if the silent label condition is not satisfied in step S502, it is determined in step S504 whether the feature amount read in step S501 satisfies a predetermined unvoiced consonant label condition. Here, the unvoiced consonant label conditions are: (ConHEnergyLow <AveEnergy <ConHEnergyMax)
In addition, ・ (AveZeroCrossRate> ConHZeroCrossRateLow). Then, in step S505, when the unvoiced consonant label condition is satisfied, the unvoiced consonant label is associated with the small segment of interest and stored.

In step S506, since the feature amount read in step S501 does not satisfy the silent label condition and the unvoiced consonant label condition described above, the voice pitch is tried to be detected. If the voice pitch is detected, the voice pitch is detected. A label is given to the corresponding small segment group (step S50).
7). The pitch detection will be described later in detail.

Here, the reason why the voice pitch label is given to the small segment group is that there is a possibility that the small segments may be integrated in the pitch detection described later. In that case,
This is because a plurality of small segments after the focused small segment are integrated into one in step S508 and a pitch label is given to this. At this time, the segment in which the voice pitch is detected is a vowel mainly accompanied by vocal cord vibration.

If the voice pitch cannot be detected in step S506, the voiced consonant label condition determination is performed in step S509. At this time, the voiced consonant label conditions are: (ConLEnergyLow <AveEnergy <ConLEnergyMax)
In addition, ・ (AveZeroCrossRate <ConLZeroCrossRateMax). Then, in step S510, when the above voiced consonant label condition is satisfied, the voiced consonant label is associated with the small segment of interest and stored.

In step S511, since the above-mentioned conditions are not satisfied, the noise label is associated with the small segment of interest and stored.

Now, the processing steps from segmentation of the audio signal waveform to labeling will be described with reference to the example shown in FIG.

FIG. 6 is a diagram for explaining the processing steps from segmentation of the audio signal waveform to labeling in this embodiment.

More specifically, FIG. 6A shows the audio signal waveform after the low-pass filter. FIG. 6B is the same as FIG.
The state shown in (a) is divided into small segments based on the zero-crossing points of the audio signal waveform, and the thick vertical line in the figure represents the division of small segments.

FIG. 6 (c) shows the result of the audio labeling and segmentation. The slender vertical lines shown in FIG. 6 represent segment breaks, and the thick vertical lines show the remnants of the integrated small segments. Is shown. In FIG. 6 (c), it can be seen that some of the plurality of small segments divided as shown in FIG. 6 (b) are integrated into one pitch segment, and the assigned label is attached to each segment. It is shown.

(Voice Pitch Detection) Next, the operation of the voice pitch detector 25 will be described with reference to FIGS. 9 and 10. This process corresponds to step S303 in FIG.

FIG. 9 is a flow chart showing the voice pitch detection processing in this embodiment.
25 shows a processing procedure performed by 25.

In step S901 shown in the figure, the zero crossing point information of the audio signal waveform after the low pass filter is obtained. Then, the voice pitch is obtained by verifying the similarity of the waveforms with the zero crossing point as a reference.

FIG. 7 is a diagram exemplifying a voice signal waveform for explaining the voice pitch detection processing in this embodiment.

In this embodiment, the reference zero-crossing point is the start point of the waveform having a positive value in the time direction,
In the example of FIG. 7, the reference zero crossing points are X1, X2, and X3.

Then, in step SS902, in the case illustrated in FIG. 7, the zero crossing point X1 is set as the starting point and the zero crossing point is set.
F (x) is the partial waveform with X2 as the end point, g (x) is the partial waveform with the zero crossing point X2 as the starting point and the zero crossing point X3 as the ending point.
Determine as an initial standard.

Then, in step S903, it is determined whether or not there is an unprocessed voice section (voice segment). If it exists, the process proceeds to step S904, and if it does not exist, the process ends.

In step S904, pitch extraction processing for reporting the presence / absence of a voice pitch and its segment range is performed. Here, the timing to report is the timing at which the voice pitch segment is interrupted, or the partial waveform f (x).
This is the case when the pitch for is not found. The pitch extraction process in step S904 will be described with reference to FIG.
Details will be described later with reference to 0.

Then, in step S905, it is determined whether or not there is a voice pitch.
In step S906, the voice pitch segment information is
It is stored in association with the voice section (voice segment) of interest. On the other hand, if there is no voice pitch, the process returns to step S903.

Here, the pitch extraction processing performed in step S904 will be described in detail with reference to FIG.

FIG. 10 is a flowchart showing the voice pitch detection processing in this embodiment, in which step S904 is executed.
It is a flowchart which shows the detail of a process of (FIG. 9).

In the figure, in step S1001, g (x) is set for the set f (x). And
In step S1002, the set length of f (x) is checked, and if it is too long to exist as a pitch,
It is determined that there is no voice pitch corresponding to the f (x), and in step S1003, the end point of the f (x) is set as the starting point,
Of the zero-crossing points that are the end points of the waveform with a negative value in the time direction, a new partial speech segment f (x) whose end point is the closest one to the start point has been set, and attention has been paid until now.
Report the segment of f (x) as not a pitch segment.

Furthermore, in step S1004, the focused f
Check the length of (x), and if it is too short to exist as a pitch, pay attention in step S1005.
Of the zero crossings that have the end point of f (x) as the start point and the end point of the waveform with a negative value when viewed in the time direction, the one closest to the start point (end point of f (x)) is the end point. By integrating the partial voice segment which has as the above at the end of the focused f (x), the process returns to step S1001 as a new f (x).

On the other hand, in step S1006, step S10
The dissimilarity calculation with g (x) is performed on the focused f (x) that has passed the check in 02 and step S1004. The dissimilarity calculation performed in this step is calculated using the following dissimilarity evaluation function.

That is, the absolute value of the difference between f (x) and g (x) at time Xf of the partial voice segment f (x) is Δ.
(Xf), X1 ≤ Xf ≤ X2 and Xg = X2 +
As (Xf-X1), it is represented by Δ (Xf) = | f (Xf) -g (Xg) |. Even in this case, f (x) and g (x)
Based on the square of the difference, not the absolute value of the difference, Δ (Xf) = [f (Xf) −g (Xg)] × [f (Xf
) -G (Xg)].

And further,

[0102]

[Equation 2] It can be expressed as.

Then, in step S1007, it is determined whether the dissimilarity calculated as described above is equal to or greater than the threshold value ETh. If DiffSum ≧ ETh, the process returns to step S1005. If DiffSum <ETh, the process proceeds to step S1005. In order to perform a precise voice pitch detection, in step S1008, the positions of f (x) and g (x) are corrected so that the small section having the highest energy is at the end of the pitch segment.

FIG. 8 is a view for explaining the pitch detection reference update procedure performed in the voice pitch detection processing in this embodiment. Correcting the pitch reference in the small section having the largest energy is also rational because the small section is a waveform generated immediately after the vocal cord vibration.

Next, in step S1009, the pitch detection counter is reset to 0, in step S1010, dissimilarity calculation is performed in the same manner as in step S1006 described above, and in step S1011, the calculated dissimilarity and threshold ETh are set. The comparison process is performed in the same manner as step S1007 described above.

Then, as a result of the comparison in step S1011, if the calculated dissimilarity is greater than or equal to the threshold value ETh, the process proceeds to step S1013, and if the dissimilarity degree is less than the threshold value ETh, the process proceeds to step S1014. .

In step S1013, it is determined whether the voice pitch is detected twice or more. If it is less than two times, the voice segments are integrated in step S1005 described above. Since it can be determined that the voice pitch segment is detected, in step S1015, the end point of g (x) is set as the start point, and the zero crossing point closest to the start point that is the end point of the waveform having a negative value in the time direction is set as the end point. A new segment f (x) is set, and the pitch segment range indicating that the pitch segment is detected is reported.

In step S1014, the pitch detection count is incremented, and the current end point of g (x) is set as the start point.
Among the end points of the waveform having a negative value in the time direction, a new partial voice segment f (x) having the nearest zero-crossing point as the end point is set to the start point and the partial voice segment f (x) is set. A new g with the zero crossing point that is the closest to the point that is the end point of the waveform having a negative value in the time direction as the end point.
(X) is set, and the process returns to step S1010.

The voice pitch segment obtained by the voice pitch detection process (FIGS. 9 and 10) described above is stored in a memory (not shown) for use by the voice section determination unit 28 in the subsequent stage.

(Voice section determination) Next, the voice section determination unit 28
Then, the voice segment of the person (section A) is determined using the voice pitch segment acquired by the above voice pitch detection processing. This process corresponds to step S304 in FIG.

Generally, in the case of a pure human voice, the vowels occupy most of its voice section, and therefore the segment in which the pitch exists appears stably for a long time. On the other hand, it has been empirically known that when there is BGM, it is affected by the temperament, but when human voice energy is somewhat higher than the energy of BGM, it is not so affected. In addition, the voice energy is GB in a certain partial section.
If it is not sufficiently larger than the energy of M, the exact pitch does not appear in that subsection.

In many cases, consonants are accompanied immediately before a vowel, but no pitch appears even in the case of a consonant that does not accompany the vibration of the vocal cords, and the duration is a short burst with a duration of 10 ms or less. Yes, even the longest fricative is on the order of tens of ms. In addition, silence may occur immediately before a plosive sound is generated.

Therefore, the segment for which the voice pitch is obtained becomes discrete not only due to the factors external to the apparatus but also due to the factor of the human voice itself. Even in such a case,
It is necessary to determine the human voice section (section A) by integrating the calculation results of the voice pitch cycles of the partial sections by considering the front and back or the entire pitch cycle and further utilizing the characteristics of the voice.

FIG. 11 is a flow chart showing the voice section determination processing in this embodiment.
The processing procedure performed by is shown.

In the figure, first, in step S1101, a continuous silence, unvoiced consonant label, voiced consonant label,
Alternatively, a group of segments having noise labels is combined into one segment.

Further, in step S1102, continuous pitch label segments are obtained and combined to obtain
The average pitch period of those multiple segments is calculated. This integrated pitch segment is called "integrated pitch segment".
Will be called.

In step S1103, a segment associated with noise labels sandwiched between integrated pitch segments and associated with a noise label is obtained. In step S1104, the average pitch period variation rate of integrated pitch segments at both ends of the segment is a threshold value. It is determined whether or not the value is Th1 or less, and if this condition is satisfied, these segments are integrated into one integrated pitch segment in step S1105. By this processing, it is possible to correct even if the BGM having large energy overlaps the pitch segment, that is, a part of the vowel.

By the way, in most cases, since a single consonant does not exist, a consonant usually accompanies backward or forward. This is called a CVC (Consonant Vowel Consonant) model.

Therefore, in step S1106, the unvoiced consonant segment, the voiced consonant segment, and the pitch segment are integrated on the basis of this CVC model to obtain a voice section. Here, details of the processing in step S1106 will be described with reference to FIG.
Will be described with reference to.

FIG. 12 is a step S1106 in the flowchart showing the voice section determination processing in this embodiment.
It is a flowchart which shows the detail of a process of (FIG. 11).

In the figure, in step S1201, the leading integrated pitch segment is set as a reference integrated pitch segment. Next, in step S1202, an integrated pitch segment next to the reference integrated pitch segment is obtained.

Further, in step S1203, it is judged whether or not a voiced consonant segment or an unvoiced consonant segment exists between the two integrated pitch segments, and if there is not, the next integrated pitch segment as the reference in step S1206 is integrated. It is determined whether or not a pitch segment exists, and if it does not exist, the process ends. If it does exist, the integrated pitch segment serving as a reference is set in step S12.
Update at 07.

On the other hand, if it is determined in step S1203 that a voiced consonant segment or an unvoiced consonant segment exists between the two integrated pitch segments, the distance Dist between the two integrated pitch segments is the threshold Pima.
In step S1204, it is determined whether or not x1 or less.
If the distance Dist is less than or equal to the threshold value Pimax1, in step S1205, the end points of the two integrated pitch segments are stored as the voice section of the person having the end point and the start point.

Here, as the threshold value Pimax1, it is preferable to use a consonant having a normal longest duration, for example, a duration sufficiently longer than the duration of an unvoiced fricative / S / etc., in which case 2
There may be silent segments as well as consonant segments between two integrated pitch segments. The reason is,
This is because, among unvoiced consonants, a plosive sound or an affricate sound may cause a short silence before vocalization.

After the voice section is stored in step S1205, in step S1206, it is determined whether or not there is an integrated pitch segment next to the reference integrated pitch segment. If it does not exist, the process is terminated, and if it exists, it is determined. Updates the reference integrated pitch segment in step S1207, and repeats the processing of each step described above until the ending condition of step S1206 is satisfied. However,
The integrated pitch segment information and its average pitch information are saved without being discarded for the next processing.

On the other hand, as a result of comparing the average pitch periods of the two integrated pitch segments in step S1204, when the period variation rate is larger than a certain threshold value Pimax1, the above-mentioned steps S1206 and subsequent steps are performed.

Now, let us return to the description of the flowchart of FIG. 11 again. In step S1107, the CVC structure is not taken,
Considering the case of a VV (Vowel-Vowel) structure such as “blue”, a voice segment is integrated by integrating two pitch segments having a silent segment or a noise segment between adjacent segments based on the VV model. Ask.

Here, the voice section detection processing performed in step S1107 will be described in detail with reference to FIG.

FIG. 13 is a step S1107 of the flowchart showing the voice section determination processing in this embodiment.
It is a flowchart which shows the detail of a process of (FIG. 11).

In the figure, in step S1301, the first integrated pitch segment is set as a reference integrated pitch segment. Next, in step S1302, an integrated pitch segment next to the reference integrated pitch segment is obtained.

Further, in step S1303, it is determined whether the distance Dist between the two integrated pitch segments is less than or equal to a certain threshold Pimax2. If the distance Dist is greater than the threshold Pimax2, the process proceeds to step S1306, and the sensory Dist is detected. Is the threshold Pi
If max2 or less, the process advances to step S1304.

In step S1304, there is a threshold Th2 having the average pitch period variation rate of the two integrated pitch segments.
In the following cases, in step S1305, the segment sandwiched between the two integrated pitch segments is stored as a voice section. In that case, a silence segment or a noise segment may exist between the two integrated pitch segments in order to increase resistance to disturbance.

After storing the voice section in step S1305, in step S1306, it is determined whether or not there is an integrated pitch segment next to the reference integrated pitch segment. In this case, the reference integrated pitch segment is updated in step S1307, and the process is repeated until the ending condition of step S1306 is satisfied.

On the other hand, as a result of comparing the average pitch periods of the two integrated pitch segments in step S1304, when the period variation rate is larger than the threshold value Th2, the process proceeds to step S1306 described above and the same process is performed.

Even if a noise label is generated due to BGM or the like contained in the voice with reference to the segment for which the pitch is detected in this way, in the present embodiment, the noise label is added in the above-mentioned voice section determination process. Even in this case, the integrated pitch segments are integrated by considering the continuity of the average pitch period of the integrated pitch segments before and after the integration, and the unvoiced consonant segment and the voiced consonant segment are separated by introducing the CVC model. It is possible to extract a voice segment that is strong against external disturbances by using the features of the voice, by determining the voice segment by combining the integrated pitch segments that exist in the Becomes

(Correction of human voice section) As described above,
In the voice section correction process performed after detecting the voice section (section A) of the person, the person is located in the vicinity on the time axis so that the person does not feel uncomfortable when listening to the reproduced voice based on the processing result. Correction is performed by integrating a plurality of voice sections as one voice section. The reason is that, for example, when the interval between two sections A located near each other on the time axis is narrow, the double speed reproduction accompanied by voice is performed at a speed at which a person can understand the content by listening to the section A, while the section B is reproduced. On the other hand, if reproduction is performed at a high speed with a high magnification within a range in which a person can understand the content by viewing the reproduced video, the reproduction mode is greatly changed, and it is difficult for the user to hear.

Also, from the viewpoint of the moving picture decoder and the reproduction processing, a change in speed in a short section has a large processing overhead, the reproduction operation is temporarily stopped, and the reproduction becomes jerky. As a result, the same phenomenon has been observed in many other moving image reproducing means in addition to the observation in the experiment by the applicant of the present invention using Microsoft DirectShow.

Therefore, in the present embodiment, when the interval between the two voice sections (section A) located closest to each other on the time axis is equal to or less than a certain threshold (Th3 in FIG. 14), these voice sections are set. Correction is made by integrating the sections. In determining the threshold value, for example, a scene in which a conversation is conducted is assumed, and the degree to which the conversation is established is experimentally obtained and used as the threshold value. The procedure of the processing in this case will be described with reference to FIG.

FIG. 14 is a flow chart showing an integrated correction process performed for a voice section having a short interval in this embodiment. This process is a process performed by the voice section determination unit 28, and represents the details of the voice section correction process (step S103) described above.

In step S1401, the first detected section A on the time axis among the plurality of sections A detected earlier is read as the target speech section, but there is no target speech section. For example, this process ends.

In step S1402, it is determined whether or not there is a next speech section (section A) of interest. If there is no speech section of interest, this process is terminated. Steps S1403 to S14 to be described
The process of 07 is repeated.

In step S1403, since it is determined in step S1402 that the next voice section of interest exists, the voice section information representing the voice section (section A) is read. Here, the voice section information is information in which the start point and the end point of the voice section are paired.

In step S1404, the distance between the two sections A, that is, the distance between the end point of the previous voice section (the current voice section of interest) and the start point of the next voice section on the time axis ( Time interval), and this distance is a predetermined threshold Th
It is judged whether it is 3 or less.

In step S1405, since it is determined in step S1402 that the interval between the two sections A is less than or equal to the predetermined threshold Th3, these two speech sections are integrated into one speech section. More specifically, in the voice section information of the integrated voice section, the start point of the previous voice section is set and the end point of the next voice section is set by the process in this step.

In step S1406, the integrated voice section is set as the current voice section (section A), and the process returns to step S1402.

In step S1407, since it is determined in step S1402 that the interval between the two sections A is larger than the predetermined threshold value Th3, the current voice section is stored as it is as one corrected voice section information. At the same time, in step S1408, the next voice section is set as the voice section to be focused on as the processing target, and the process returns to step S1402.

Such integration processing is repeated until there is no speech section (section A) to be handled.

(Correction of Human Voice Section Using Scene Change Point Information) Generally, when there is a video transition in the video data including the audio signal and the section A starts immediately after that, the video is reproduced. , After the video of the beginning part of the scene is played back at a high speed for a moment, the playback video by the double speed playback accompanied by the audio is performed at a speed that can be heard and grasped by the person, and the user feels that the video is flickering. Occurs.

Therefore, in the present embodiment, for example, by adopting the scene change point detection technology disclosed in the above-cited Japanese Patent Application Laid-Open No. 2000-235639 by the applicant of the present application, among the detected scene change point groups, If there is a scene change point that is earlier in time than the beginning of the voice section after the voice section correction processing and is closest to the start of the voice section, the beginning of the voice section is By performing the correction for replacing with the information corresponding to the scene change point, the user's uncomfortable feeling at the time of the quick-view reproduction is removed. At this time, the threshold value for the proximity determination is a value corresponding to the overhead when shifting from the high-speed reproduction state to the double-speed reproduction with voice at a speed at which a person can hear and understand the contents.

FIG. 15 is a flow chart showing an integrated correction process of a voice section performed using a scene change point in this embodiment. This process is performed by the voice section determination unit 28.
The details of the above-described fast-viewing reproduction section correction process (step S104), which is the process performed in step S4.

In the figure, first, in step S1501,
The first scene change point (A) on the time axis is read from the scene change point group (scene change point information or scene change position information) detected in the scene change point detection processing (step S106).

The scene change point information is usually described in frame units, but in this step, it is converted into time information based on the frame rate and then compared with the voice section information. That is, in the algorithm of this embodiment, two consecutive scene change point information are used in order to find the closest scene change point from the start point of the voice section. Here, for convenience of explanation, the previous scene change point is used. The point is A and the next scene change point is B,
In step S1501, the time of the scene change point is stored in A direction.

In step S1502, it is determined whether or not there is unread voice section information. If there is not read voice section information, the process ends. If there is unread voice section information, one voice section information is read in step S1503.

In step S1504, it is determined whether or not there is scene change point information that has not been read. If there is no scene change point information, the voice section information that has already been read in step S1503 is directly corrected in step S1505. It is updated and stored as information.

In step S1506, since it is determined that there is scene change point information that was not read in step S1504, that scene change point information is read as scene change point information B.

At step S1507, the scene change point A
On the time axis, it is determined whether or not it is located before the start point of the current voice segment read in step S1503, and if it is, it is determined in step S1505 that no correction is necessary and the voice segment The information is updated and stored as it is as the corrected voice section information.

In step S1508, since it is determined in step S1507 that the scene change point A is located before the start point of the current voice section of interest, the scene change point A is within the threshold point Th4 and the start point of the voice section. If it is not within the threshold value Th4, it is judged whether or not there is a distance, and in step S1509, the information of the scene change point B is copied to the scene change point A, so that the next scene change point can be obtained. Prepare for judgment.

In step S1510, since it is determined in step S1508 that the scene change point A exists at a distance within the threshold Th4 from the start point of the current voice section of interest, the scene change point B is the start point of the voice section. It is determined whether or not it is located behind, and if it is not located behind, the process proceeds to step S1509.

On the other hand, when it is determined in step S1510 that the scene change point B is located behind the start point of the voice section, the scene change point A is the start point in step S1511, and The partial section whose end point is the end point is updated and stored as corrected voice section information, and in step S1512, the information of the scene change point B is copied to the scene change point A to determine the next scene change point. And prepare to.

That is, the scene change point A is located in front of the start point of the current voice segment of interest and is in the vicinity of the threshold value Th4 or less as a result of the judgments in steps S1507, S1508, and S1510 described above. , S1511 and S1512 described above are not confirmed until the point is closest to the start point of the voice section.
Is processed.

If it is determined in step S1510 that the scene change point B is not behind the start point of the voice section, the scene change point B has been corrected more than the currently set scene change point A. Since it can be determined that it is more suitable as a start point candidate of the voice section, step S
In 1509, the information of the scene change point B is copied as a new scene change point A to prepare for the determination of the next scene change point, and then the process returns to step S1504. However, since the scene change point A in this case has already satisfied the requirements of step S1507 and step S1508, step S1507
And step S1508 may be skipped and the determination in step S1510 may be suddenly made.

Speech section integrated correction processing described above (FIG. 15)
The corrected voice section information acquired by the procedure of (1) is stored in the moving picture quick reference index storage unit 11 as the quick view reproduction section information in a schema as illustrated in Table 1.

Table 1 is a table exemplifying the result of scene change detection in the present embodiment. As an example, the frames in which the scene change points are detected are converted into seconds based on the frame rate (30 images / Sec). The result is stored.

[0164]

[Table 1]

Next, Table 2 is a table exemplifying the detection result of the voice section in the present embodiment, and one voice section is represented by a start point and an end point.

[0166]

[Table 2]

Table 3 is a table exemplifying the corrected voice section detection result in this embodiment. Based on the result shown in Table 1 and the result shown in Table 2, the voice using the scene change point is shown. The processing result when the section integrated correction processing (FIG. 15) is performed with the threshold value Th4 = 2000 mSec is shown.

[0168]

[Table 3]

Referring to Tables 1 and 2, for the voice section 0 and the voice section 2, the start points of the respective voice sections are 60000 mSec and 400000 mSec, and the threshold Th4 is set.
There are no scene changes within 2000 mSec. Also, for voice section 1, the start point is 102000 m
Before 1500 mSec of Sec and within 2000 mSec, as a scene change point, scene change ID = 2 (start time
100000 mSec) and scene change ID = 3 (start time 1
01000 mSec) exists, but since the closest one is selected according to the algorithm shown in FIG. 15, as a result, 101000 mSec with scene change ID = 3 is selected, and this is reflected in Table 3.

<Movie fast view reproduction unit 200> Movie fast view reproduction process (step S10)
7) is for the human voice section (section A), while performing double speed reproduction accompanied by voice at a speed at which a person can hear and understand the content, for the section (section B) which is not the human voice section. Is a high-speed, high-speed reproduction within a range where a person can see the reproduced video and understand the content.

In recent years, a moving image reproducing environment has been prepared. For example, if a DirectShow module manufactured by Microsoft Corporation is used, it is possible to continuously reproduce by designating a speed in an arbitrary section.
By using a module having such a function, it is possible to relatively easily change the reproduction speed in an arbitrary section. At that time, what is important is the viewpoint of changing the speed.

FIG. 16 is a flow chart showing the moving image quick view reproducing process according to the present embodiment.

In FIG. 18, in step S1701, the user selects a desired one from the above-described user profiles 14, and the specific procedure is, for example, shown in FIG. A display screen including such a user profile list may be displayed, and the user may select a desired profile from the display screen by using a remote control terminal or the like.

That is, in order to specify the profile desired by the user in the user profile list shown in FIG. 18, for example, an operation button for profile selection is provided on the remote control terminal, and when the user presses this operation button, the operation shown in FIG.
A menu display screen as illustrated in 0 is displayed, and the user designates a desired profile by using the operation button for profile selection of the remote control terminal while watching the screen. Of course, to select the user profile,
An automatic profile selection method that uses personal recognition technology such as fingerprints, voiceprints, face recognition, etc. is also conceivable.Since this method can always specify the correct profile, it may cause an error in specifying the profile or the profile of another person. You can prevent troubles such as changing and looking at the contents.

When a user profile is newly registered, "new registration" is displayed on the display screen of FIG.
When the button is specified with the pointer device, the display screen shown in FIG. 19 for inputting the profile name and other attributes appears.

That is, FIG. 19 is a view showing an example of a display screen for user profile registration. In the initial state, the contents other than the identification name and age are filled with the reference value, and the user's input operation makes it unique. Only the place where the identification name and age need to be changed is changed, and after passing the appropriate check of the predetermined input value range, the profile is changed to the user's profile when the user presses the “OK” button. It is newly additionally registered in the profile 14.

When the user desires to change the contents of a desired profile, the "change" button is pressed on the display screen shown in FIG. 18, and the desired profile is selected on the display screen shown in FIG. On the displayed display screen of FIG. 19, after changing the information content of the item desired to be changed, the “OK” button may be pressed.

Further, when the user desires to delete the desired profile, the "delete" button is pressed on the display screen shown in FIG. 18, the desired profile is selected on the display screen shown in FIG. 20, and then "OK". ] Button.

When the "Cancel" button is pressed on the display screens shown in FIGS. 18 and 19, the processing (registration, change, deletion of the profile) corresponding to the selection operation or the input operation up to that point is performed. ) Is not performed and the process ends.

Next, in step S1702, step S17
It is determined whether the profile selected in 01 exists in the user profile 14, and if it exists, the target profile is selected in step S1703.
Read from profile 14 and if not present,
The playback speed of the section A and the section B, which is preset as a reference value, and the volume of the section B during playback,
It is read in step S1706. Here, an example of the data schema of the user profile is shown in Table 4.

[0181]

[Table 4]

Table 4 is a table exemplifying the user profile in this embodiment. The reference value may be stored as shown in profile ID = 0. In this case,
The playback speed in section A is 1.5 times faster, and the playback speed in section B is 10.0.
The reference value of the volume at the double speed and the reproduction of the section B is 0 (that is, the audio mute). This value is used as the reference value used when the user profile is newly registered.

Further, in the data schema 1 of the user profile in Table 4, “None” means that no value has been set. Conversely, when a value has been set, that value is given the highest priority for playback. To do. Further, in Table 4, Good and Poor in the columns of visual acuity and hearing acuity represent the ability of the visual acuity and the hearing ability of fast voice, which are irrelevant to the age of the person.

In general, the older the person becomes, the harder they are to hear, and the slower the speed of understanding words is. In many cases, the child can understand that he / she reproduces voice at a high speed because his / her language ability is underdeveloped. It often disappears.

Based on these circumstances, a template of the playback speed of the section A and the playback speed of the section B suitable for the age of the healthy person is prepared in advance, and based on the age stored in the user profile 14, Determine these speeds.

However, even if a young man has weak visual acuity and hearing of fast voice, or if he / she is a foreigner and the language is different from his native language (eg Japanese), the voice cannot be understood at a high speed. , There may be age-independent causes. Therefore, in the present embodiment, as in the user profile illustrated in Table 4, the characteristics of the visual acuity and the hearing ability are described, and if there are these settings, this is prioritized.
The reproduction speed of the section A and the reproduction speed of the section B are determined to be low.

In such a case, for the elderly and users with weak visual acuity, the reproduction speed in the human voice section (section A) may be slower than the normal speed, although this may deviate from the viewpoint of the original quick-view reproduction. By determining the speed and setting the playback speed of the section (section B) that is not the human voice section to be equal to or higher than the normal speed, the user as a whole can perform low-speed playback while grasping the voice content of the section A, but as a whole. Makes it possible to view videos in a faster time than in slow playback of all sections.

For users who have a weak hearing ability for fast-talking voices and users who cannot understand in fast-talking Japanese due to foreigners, the playback speed of section A is determined to be slower than the normal speed and section B is set. Regarding the playback speed of
By setting the playback speed to be the same as that of a healthy person of that age, while performing low-speed playback capable of grasping the audio content of section A,
As a whole, it is possible to view a moving image in a faster time than in slow-moving all sections.

As described above, in the present embodiment, the speed determination process for the user profile is performed in advance by the template of the reproduction speed of the section A and the reproduction speed of the section B suitable for the age of a healthy person, the visual acuity and the hearing ability of fast voice. We make a comprehensive judgment by taking into consideration the weak symptoms of Japanese and the situation that we cannot catch up with fast-paced Japanese because of foreigners.

Further, in the present embodiment, the determination as to whether or not the user is fluent in the language of the voice content is whether or not the user is proficient stored in the user profile 14, or language type information for specifying the native language, This is performed by comparing the language type information of the audio content included in the moving image to be reproduced. recent years,
Digital content such as a DVD and digital media such as a digital BS store language type information for specifying the language of audio content, and in recent years, program content can be electronically obtained from EPG (electronic program guide) or the like. Therefore, it is realistic to use such information. Even if this information is not available, even in the case of terrestrial TV programs, by default, the main language is the native language and the sub-audio is the foreign language in the two-language audio. It may be estimated based on the rule.

In step S1704, the reproduction speed of section A and the reproduction speed of section B are determined based on the profile desired by the user and read in step S1703. Here, details of the processing in this step will be described with reference to FIG.

FIG. 17 is a step S1704 in the flowchart showing the moving image fast-viewing reproduction process according to this embodiment.
17 is a flowchart showing details of the process of (FIG. 16).

In the figure, first, in step S1801,
The profile previously selected by the user is read from the user profile 14, and in step S1802,
According to the age of the user acquired from the read profile, by referring to the template in which the reproduction speed of the optimum section A and the reproduction speed of the section B are set according to the age of the healthy person, the section A for the user is referred to. And the playback speed of the section B are provisionally determined.

In step S1803, it is determined whether or not the profile read in step S1801 describes that the dynamic visual acuity is weak.
In step S1804, the reproduction speed of the section A and the section B
Both the playback speed and the playback speed are updated to values lower than the reference value.
Therefore, this value is also preferably stored in the profile in advance.

[0195] In step S1805, it is determined in step S1803 that the profile does not describe that the dynamic visual acuity is weak. Therefore, it is determined whether the profile describes that the hearing ability of fast voice is weak. If it is described, in step S1806, the section A
Only the playback speed of is updated to a lower value. Therefore, this value is also
It is desirable to store it in the profile in advance.

[0196] In step S1807, since it is determined in step S1805 that the hearing ability of fast audio is weak, it is possible to obtain the language type information of the audio content included in the moving image data to be reproduced. If it is available, the process proceeds to step S1808,
If it is not available, the process ends.

In step S1808, the language type information of the audio contents included in the moving image data to be reproduced is obtained, and the obtained language type information is compared with the favorite language information described in the currently selected profile. If these two types of information match, the process is terminated. If they do not match, only the reproduction speed of the section A is updated to a low value in step S1809. Therefore, this value is also preferably stored in the profile in advance.

That is, in the series of processing shown in FIG. 17, if none of step S1803, step S1805, and step S1808 is met, the reproduction speed of the section A temporarily determined in step S1802 and the section B The playback speed will be adopted as it is.

If the dynamic visual acuity and the hearing ability of fast voice are excellent in spite of being old or young, or if they are inferior on the contrary, the replay speed of section A and the replay speed of section B are changed. It is preferable to use the configuration so that these values can be changed. In this case, the user appropriately changes the reproduction speed of the section A and the reproduction speed of the section B while watching the reproduced video, and automatically or after asking the user for confirmation, sets the set reproduction speed information. By storing the profile in the profile corresponding to the user, it is possible to perform quick-playback playback of the moving image that reflects the previous operation information and is easy to understand for each user.

If the above-mentioned profile is not used and is simply performed, for example, instead of the processing in each of steps S1701 to S1704 and step S1706, the reproduction speed of the section A is changed from 0.5 × to 2 ×. It is assumed that the reproduction speed of the section B can be variably set by the user using the operation menu from 2 × to 10 ×.

By the way, when the section B is reproduced at a high speed and at a high speed, a sound of "curculer" is produced. However, if the user does not want to hear the sound, the sound reproduction is muted when the section B is reproduced. It is envisioned that an embodiment may be provided in which the output is turned off or the volume is changed to a low level. Such settings are also described in advance in the profile read in step S1703, and during the quick playback of the moving image,
If such a profile is given the highest priority, and it is determined in step S1702 that no profile exists, a preset reference volume is adopted in step S1706.
Of course, if it is performed in a simpler manner, for example, an embodiment is conceivable in which the moving image quick playback process predetermines how to process the audio playback level of the section B.

With the above configuration, in the present embodiment, the reproduction speed of the section A and the reproduction speed of the section B, or both of them, and the audio level of the section B are specified by using the user profile. It is possible to easily realize the optimum reproduction for the user.

Next, in step S1705, the quick view reproduction section information which is the audio section information corrected in the quick view reproduction section correction processing (step S104) is read from the moving picture quick view index storage unit 11, and in step S1707 the section A is read. The playback time of the section A is calculated by dividing the total length of the section by the playback speed, the playback speed of the section B is calculated in the same manner, and the time required for the user to quickly look is calculated by adding these two values. . Then, the calculated time required for the quick look is presented to the user using the display 23 or the like.

In step S1708, it is determined whether the user who has recognized the fast-view playback time in step S1707 is satisfied with the time by using an input operation or the like on the remote control terminal, and the user is satisfied with this determination. If so, in step S1710, the moving image to be played back stored in the moving image data storage unit 10 in accordance with the playback speeds of the sections A and B and the audio playback level of the section B set by the above-described processing. To play.

In step S1709, since it is determined that the user is not satisfied in step S1708, the reproduction speeds of section A and section B and the sound reproduction level of section B are set so that the reproduction time desired by the user is reached. By providing a man-machine interface that can be changed, the reproduction time is adjusted so as to be close to the reproduction time desired by the user who is not satisfied with the profile and standard settings, and the process returns to step S1707.

As another embodiment corresponding to step S1709, while watching the moving image reproduction based on the currently set reproduction speeds of the section A and the section B, the reproduction speed desired by the user is set for each section. Configurable to be modifiable,
A configuration is also conceivable in which the time required for quick viewing is calculated and presented accordingly, and the reproduction time is adjusted to be closer to the reproduction time desired by the user who is not satisfied with the profile and standard settings.

[0207] By the way, regarding the relationship between the user profile and the speed instruction desired by the user, the user who saw the time required for the quick-motion playback of the moving image in step S1707 has a section in order to set the time required for the quick playback of the desired moving image. When these settings are adjusted / changed using a man-machine interface capable of changing the reproduction speeds of A and section B, it may be desirable to use the adjusted / changed value as the reference value. Therefore, in such a case, if "Yes" is selected after prompting the confirmation by the user automatically or by the confirmation screen illustrated in FIG. 21, the reproduction speed adjusted / changed by the user. By storing the information in the profile corresponding to the user, it is possible to perform easy-to-understand movie quick-view playback according to the user while reflecting the previous operation information in the subsequent moving image playback.

In the above-described embodiment, the number of zero crossings and the voice energy are used as the voice labeling process, but the specific processing procedure is not necessarily limited to the above algorithm, and a known feature amount may be used. Alternatively, different label determination algorithms may be used.

That is, the purpose of the voice detection processing according to the above-mentioned embodiment is to use the zero-crossing information of the low-pass filtered voice signal to convert the voice signal into a plurality of reasonable voice segments (voice sections). Of the CVC voice model, etc., based on a predetermined voice pitch that is always accompanied by a vowel that occupies most of the human voice after the voice pitch is detected by waveform processing and the voice labeling is performed. Even if a disturbance such as BGM is included in the audio signal, the process of recovering the disturbance is included by integrating the plurality of audio segments by using.

Therefore, there are no restrictions on how to implement the AGC 21 and the low-pass filter 22, and the speech labeling is not necessarily restricted to the algorithm of this embodiment, and different label determination algorithms may be used.

Also, in the determination processing (FIG. 11) performed by the voice segment determining unit 28, the voice segment is integrated by integrating the unvoiced consonant segment or the voiced consonant segment and the pitch segment, which is performed in step S1106. The order of the process of obtaining and the process of obtaining the voice section by integrating two pitch segments having a silent segment or a noise segment, which is performed in S1107, is limited to the above-described embodiment. Instead, an algorithm that processes these processes in parallel may be used.

Further, in the above-described embodiment, as a procedure for selecting a user profile, the user appropriately designates the profile selection screen by using the remote control terminal, and the user profile list displayed on the display 12 is displayed. The example of the configuration in which one's user profile is selected from the above is explained, but the configuration is not limited to this. For example, even if a password is used to prevent operations such as changing or deleting another user's user profile, good.

Further, an automatic profile selection method using a personal recognition technique such as a fingerprint, a voiceprint, or face recognition is naturally conceivable. In these cases, a password of another user cannot be selected by a password.
It is convenient because there is no need to prevent operations such as changing or deleting profiles.

In the above-described embodiment, after the user confirms the calculated time required for the quick-view reproduction, the reproduction speed of the section A and the reproduction speed of the section B are changed so as to be within the reproduction time desired by the user. By doing so, a configuration example in which a user who is not satisfied with the profile and standard settings adjusts the playback time to be close to the desired playback time,
The present invention is not limited to this configuration. For example, while the user is watching the playback video, the playback speed of the section A and the playback speed of the section B are configured to be changeable, and the time required for the quick viewing according to the setting is set. There is also an embodiment in which the reproduction time is recalculated and presented to the user to adjust the reproduction time close to the reproduction time desired by the user.

As described above, according to the present embodiment, the basis of the voice utterance mechanism uttered by a person is the vibration of the vocal cord, so-called voice pitch, and a useful voice section is obtained by extracting this from the voice signal. By detecting the true human voice section, and using that section, the synchronization between the video and the voice is not broken, and all the voices uttered by the person are played back at a speed that allows the user to grasp the content during the quick playback of the video. On the other hand, the section (section B) that does not include the voice uttered by a person is reproduced at a higher speed. As a result, it is possible to rationally reduce the total browsing time during the quick playback of the moving image, as compared with the case of performing the normal-size playback.

Further, according to the present embodiment, the reproduction speed of the section A and the reproduction speed of the section B can be easily set to the reproduction speed suitable for each user by using the user profile 14. The volume during playback of section B can also be set to be suitable for the user.

Further, according to the present embodiment, the time required for the quick-view reproduction is displayed in advance or during the reproduction of the moving image, so that the user who is not satisfied with this can display the reproduction speed of the section A and the reproduction speed of the section B. By specifying it, it is possible to adjust the time required for the quick view playback that is optimal for the user, and the information set by the adjustment can be updated and stored in the profile corresponding to the user, so the next quick view An appropriate moving image can be reproduced at the time of reproduction.

[0218]

Other Embodiments The present invention described by taking the above-described embodiments as examples may be applied to a system including a plurality of devices or may be applied to an apparatus including one device. .

The present invention supplies the software program for realizing the functions of the flow charts described in the above-described embodiments directly or remotely to the system or apparatus operating as the above-mentioned moving picture reproducing apparatus, and the system or It also includes the case where it is achieved by the computer of the apparatus reading and executing the supplied program code. In that case, the form need not be a program as long as it has the functions of the program.

Therefore, the program code itself installed in the computer to implement the functional processing of the present invention by the computer also implements the present invention. That is, the claims of the present invention include the computer program itself for realizing the functional processing of the present invention.

In this case, the program may take any form such as an object code, a program executed by an interpreter, or script data supplied to an OS as long as it has the function of the program.

As a recording medium for supplying the program, for example, a floppy (registered trademark) disk, a hard disk, an optical disk, a magneto-optical disk, an MO, a CD
-ROM, CD-R, CD-RW, magnetic tape, non-volatile memory card, ROM, DVD (DVD-ROM,
DVD-R).

In addition, as a method of supplying the program,
It can also be supplied by connecting to a homepage on the Internet using a browser of a client computer, and downloading the computer program itself of the present invention or a compressed file having an automatic installation function from the homepage to a recording medium such as a hard disk. It can also be realized by dividing the program code constituting the program of the present invention into a plurality of files and downloading each file from different homepages. In other words, a WWW (World WWW) that allows a plurality of users to download a program file for implementing the functional processing of the present invention on a computer.
An ide Web server is also included in the claims of the present invention.

Also, the program of the present invention is encrypted to C
By storing the information in a storage medium such as a D-ROM and distributing it to the user, and having the user who satisfies the predetermined conditions download the key information for decrypting the encryption from the home page via the Internet, and by using the key information It is also possible to execute the encrypted program and install the program in a computer to realize it.

Further, the functions of the above-described embodiments are realized by the computer executing the read program, and the OS and the like running on the computer execute the actual processing based on the instructions of the program. The function of the above-described embodiment can be realized also by performing a part or all of the above.

Furthermore, after the program read from the recording medium is written in the memory provided in the function expansion board inserted into the computer or the function expansion unit connected to the computer, based on the instructions of the program,
CPU provided on the function expansion board or function expansion unit
Performs a part or all of the actual processing, and the processing realizes the functions of the above-described embodiments.

[0227]

According to the present invention described above, it is possible to accurately detect a voice section uttered by a person, and to faithfully maintain a synchronization relationship between a video and a sound in accordance with the detected voice section while the user is required to browse. It is possible to provide a moving picture reproducing apparatus, a moving picture reproducing method, and a computer program therefor, which can significantly reduce the time.

[Brief description of drawings]

FIG. 1 is a diagram showing a conceptual diagram of a moving image fast viewing algorithm in a moving image reproducing apparatus according to the present embodiment.

FIG. 2 is a block diagram showing an algorithm for detecting a voice section (section A) representing a person's utterance period, which is performed in the moving image quick reference index creation unit 100.

FIG. 3 is a flowchart showing an outline of processing based on the algorithm shown in FIG.

FIG. 4 is a diagram illustrating a small segment combination process performed in the present embodiment.

FIG. 5 is a flowchart showing a voice labeling process performed in this embodiment.

FIG. 6 is a diagram illustrating a processing process from segmentation of an audio signal waveform to labeling in the present embodiment.

FIG. 7 is a diagram exemplifying a voice signal waveform for explaining a voice pitch detection process in the present embodiment.

FIG. 8 is a diagram illustrating a procedure for updating a pitch detection reference performed in a voice pitch detection process according to the present embodiment.

FIG. 9 is a flowchart showing a voice pitch detection process in this embodiment.

FIG. 10 is a flowchart showing details of the process of step S904 (FIG. 9) of the flowchart showing the voice pitch detection process in the present embodiment.

FIG. 11 is a flowchart showing a voice segment determination process in the present embodiment.

FIG. 12 is a flowchart showing details of the process of step S1106 (FIG. 11) of the flowchart showing the voice section determination process in the present embodiment.

FIG. 13 is a flowchart showing details of the process of step S1107 (FIG. 11) of the flowchart showing the voice section determination process in the present embodiment.

FIG. 14 is a flowchart showing an integrated correction process performed for a voice section having a short interval in the present embodiment.

FIG. 15 is a flowchart showing a voice section integrated correction process performed using a scene change point in the present embodiment.

FIG. 16 is a flowchart showing a moving image quick playback process according to the present embodiment.

FIG. 17 is a flowchart showing details of the process of step S1704 (FIG. 16) of the flowchart showing the moving image quick playback process in the present embodiment.

FIG. 18 is a diagram illustrating a display screen for user profile selection.

FIG. 19 is a diagram illustrating a display screen for user profile registration.

FIG. 20 is a diagram showing an example of a user profile in the present embodiment.

FIG. 21 illustrates a display screen that urges the user to confirm whether to use the adjusted / changed value as a reference value for the next and subsequent video playback when the user who is not satisfied with the time required for the video quick playback playback changes the setting. FIG.

Continuation of front page (51) Int.Cl. ⁷ Identification code FI theme code (reference) H04N 5/76

Claims (30)

[Claims] 1. A moving picture reproducing apparatus capable of reproducing moving picture information including a sound signal at a high speed, wherein a first sound section representing a human utterance period based on the sound signal contained in the moving picture information, and the first sound section. Other than the second voice section, and based on the moving image information, the first voice section performs high-speed moving image reproduction with reproduced sound at a predetermined speed at which the user can grasp the content. On the other hand, the moving picture reproducing apparatus, characterized in that the second voice section is provided with a quick-view reproducing means for performing at least high-speed moving picture reproduction at a speed higher than the predetermined speed. 2. The moving picture reproducing method according to claim 1, wherein the fast-viewing reproducing means performs moving picture reproduction at a speed higher than the predetermined speed and with at least a small volume of reproduced sound in the second sound section. apparatus. 3. The moving picture reproducing apparatus according to claim 1, wherein the quick-view reproducing means reproduces a moving picture without voice at a speed higher than the predetermined speed in the second sound section. 4. The voice section determination means extracts a voice pitch corresponding to vocal cord vibration based on the voice signal,
The moving image reproducing apparatus according to claim 1, wherein the first voice section is determined based on the extracted voice pitch. 5. The volume at the time of voice reproduction in the second voice section is determined in advance or can be designated by the user in the quick-view reproduction means.
The moving image reproducing apparatus according to claim 3. 6. The voice section determining means extracts a pitch in a possible vocal cord frequency range from a signal obtained by filtering a voice band emitted by a person included in the voice signal. 4. The moving image according to claim 1, wherein the first vowel section is determined by detecting a dominant vowel part of the voice and integrating the detected vowel parts. Playback device. 7. The voice section determining means, when determining the first voice section based on the voice signal, includes a correcting means for integrally correcting a plurality of first voice sections that are close to each other on a time axis. The moving picture reproducing apparatus according to any one of claims 1 to 3, further comprising: 8. The correcting means detects a scene change point included in the moving image information, and, in the detected individual scene change points, is earlier in time than the start point of the first voice section of interest. When the time interval between the nearest neighboring scene change point and its starting point is less than or equal to a predetermined threshold value, the starting point of the first voice section of interest is the information corresponding to the neighboring scene change point. The correction is performed by replacing
The described video playback device. 9. The quick-view reproduction means calculates a time required for the high-speed moving image reproduction based on the length of the first audio section, the reproduction speed of the section, and the length of the second audio section. The moving picture reproducing apparatus according to any one of claims 1 to 3, wherein the calculated required time is presented to the user. 10. The quick-view reproduction means, in response to the presentation of the required time, when the operation of changing the reproduction speed of the first and second voice sections is performed by the user, the reproduction after the change is performed. The moving image reproducing apparatus according to claim 9, further comprising an adjusting unit that adjusts the required time based on a speed. 11. A user profile, in which attribute information relating to each user is registered, is provided for users who can use the moving image reproduction apparatus, and the quick-view reproduction means is registered in the user profile. 4. The moving picture reproducing apparatus according to claim 1, wherein the reproducing speeds of the first and second voice sections are automatically determined according to the attribute information regarding the specific user. 12. The user profile includes at least one of age, language used, visual acuity, and fast audio hearing as attribute information regarding the individual user. The described video playback device. 13. The quick-view playback means sets the length of the first voice section, the playback speed of the section, and the length of the second voice section, which are automatically determined according to the attribute information about the specific user. Based on the calculated required time for the high-speed moving image reproduction, presenting the calculated required time to the user, and presenting the required time, the reproduction speeds of the first and second audio sections. 13. The moving image reproducing apparatus according to claim 11 or 12, further comprising: an adjusting unit that adjusts the required time based on the changed reproduction speed when the changing operation is performed by the user. 14. The adjusting means stores the changed reproduction speeds of the first and second voice sections in association with the attribute information regarding the specific user in the user profile, and the fast-view reproduction means 14. The moving picture reproducing apparatus according to claim 13, wherein the reproducing speeds of the changed first and second voice sections stored in the user profile are reflected when the high speed moving picture is reproduced. 15. The quick-view reproduction means, when the information regarding the reproduction manner of the second voice section is designated by the user, the reproduction manner with respect to the attribute information regarding the user stored in the user profile. Information relating to the reproduction mode of the second voice section, which is stored in the user profile and is stored in the user profile, and is reflected in the user profile when the high-speed moving image is reproduced. 11. The video playback device according to item 11. 16. The moving picture reproducing apparatus according to claim 1, wherein the predetermined speed during high speed moving picture reproduction in the first audio section is 1.5 to 2 times as fast as constant speed reproduction. 17. The fast-view reproduction means, when the attribute information relating to the user registered in the user profile includes identification information indicating that the user is an elderly person, a visually impaired person, or a hearing impaired person, When performing the high-speed moving image reproduction for the user corresponding to the identification information, the reproduction speed of the first voice section is slower than the normal speed and the reproduction speed of the second voice section is faster than the normal speed. 12. The moving picture reproducing apparatus according to claim 11, which is performed. 18. The fast-view reproduction means includes identification information indicating a language used by the user in the attribute information about the user registered in the user profile, and the identification information and the moving image information are included. If the language type information does not match, when the high-speed moving image reproduction is performed for the user corresponding to the identification information, the reproduction speed of the first voice section is set to be slower than the equal speed, and the second voice is reproduced. 12. The moving picture reproducing apparatus according to claim 11, wherein the reproduction speed of the section is 5 to 10 times. 19. The user profile is intended for a plurality of users who can use the video playback device.
Attribute information about each user is registered, the fast-view playback means, according to the selection operation of the specific user,
12. The moving picture reproducing apparatus according to claim 11, wherein the attribute information regarding the specific user is acquired from the user profile based on a personal authentication technique. 20. The moving picture reproducing apparatus according to claim 11, further comprising attribute information changing means capable of changing attribute information relating to a specific user registered in the user profile by the specific user himself. 21. A moving picture reproducing method for reproducing moving picture information including a sound signal at a high speed, wherein a first sound section indicating a vocalization period of a person based on a sound signal included in the moving picture information, and other than that. And a high-speed moving image reproduction with a reproduced sound at a predetermined speed at which the user can grasp the content, based on the moving image information. Then, the second voice section has a fast-viewing reproduction step of performing at least high-speed moving picture reproduction at a speed higher than the predetermined speed, the moving picture reproduction method. 22. The moving image reproduction according to claim 21, wherein, in the fast-view reproduction process, a moving image reproduction is performed at a speed higher than the predetermined speed and with a reproduction sound of at least a small volume in the second audio section. Method. 23. The moving picture reproducing method according to claim 21, wherein, in the fast-viewing reproducing step, the moving picture is reproduced in the second voice section at a speed higher than the predetermined speed and without voice. 24. In the voice section determination step, a voice pitch corresponding to vocal cord vibration is extracted based on the voice signal, and the first voice section is determined based on the extracted voice pitch. 21 to 23
The video playback method described in any one of 1. 25. In the voice segment determination step, a human voice is extracted by extracting a pitch in a possible vocal cord frequency range from a signal obtained by filtering a voice band emitted by a person included in the voice signal. 24. The moving image according to claim 21, wherein the first vowel section is determined by detecting a dominant vowel section of the voice and integrating the detected vowel sections. How to play. 26. In the voice section determination step, when the first voice section is determined based on the voice signal, a plurality of first voice sections that are close to each other on a time axis are integrally corrected. 21 to 23.
The video playback method described in any one of 1. 27. In the voice section determination step, at the time of the correction, a scene change point included in the moving image information is detected, and among the detected individual scene change points, a starting point of the first voice section of interest is detected. Also, if the time interval between the neighboring scene change point located closest in time and the closest point is less than or equal to a predetermined threshold value, the starting point of the focused first voice section is set to the neighboring scene. 27. The moving image reproducing method according to claim 26, wherein the correction is performed by replacing the information with the information corresponding to the change point. 28. Further, the method further comprises a registration step of registering attribute information regarding each user as a user profile for a user who can use the moving picture reproducing apparatus, and in the quick view reproducing step, the user profile 24. The reproduction speeds of the first and second voice sections are automatically determined according to the attribute information about the specific user registered in.
The video playback method described in any one of 1. 29. A computer program, which gives an operation instruction which can be realized by a computer, to the moving picture reproducing apparatus according to claim 1. 30. A computer program according to any one of claims 21 to 28, which gives an instruction to operate a computer.