WO2004025625A1 - 信号処理システム、信号処理装置および方法、記録媒体、並びにプログラム - Google Patents
信号処理システム、信号処理装置および方法、記録媒体、並びにプログラム Download PDFInfo
- Publication number
- WO2004025625A1 WO2004025625A1 PCT/JP2003/010824 JP0310824W WO2004025625A1 WO 2004025625 A1 WO2004025625 A1 WO 2004025625A1 JP 0310824 W JP0310824 W JP 0310824W WO 2004025625 A1 WO2004025625 A1 WO 2004025625A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- signal
- input
- frequency component
- processing
- class code
- Prior art date
Links
- 238000012545 processing Methods 0.000 title claims abstract description 204
- 238000000034 method Methods 0.000 title claims abstract description 125
- 238000001228 spectrum Methods 0.000 claims abstract description 121
- 230000005236 sound signal Effects 0.000 claims abstract description 59
- 230000008569 process Effects 0.000 claims description 101
- 238000000605 extraction Methods 0.000 claims description 22
- 230000002194 synthesizing effect Effects 0.000 claims description 15
- 238000003672 processing method Methods 0.000 claims description 12
- 230000015572 biosynthetic process Effects 0.000 claims description 8
- 238000003786 synthesis reaction Methods 0.000 claims description 8
- 239000000284 extract Substances 0.000 abstract description 6
- 230000005540 biological transmission Effects 0.000 description 50
- 238000006243 chemical reaction Methods 0.000 description 38
- 238000010586 diagram Methods 0.000 description 37
- 230000010354 integration Effects 0.000 description 17
- 238000004891 communication Methods 0.000 description 11
- 230000001413 cellular effect Effects 0.000 description 10
- 230000008878 coupling Effects 0.000 description 8
- 238000010168 coupling process Methods 0.000 description 8
- 238000005859 coupling reaction Methods 0.000 description 8
- 238000013139 quantization Methods 0.000 description 8
- 238000002360 preparation method Methods 0.000 description 7
- 238000000926 separation method Methods 0.000 description 7
- 230000006870 function Effects 0.000 description 5
- 238000005070 sampling Methods 0.000 description 4
- 230000003595 spectral effect Effects 0.000 description 4
- 230000009466 transformation Effects 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 238000009825 accumulation Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000007906 compression Methods 0.000 description 2
- 238000012937 correction Methods 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 239000004065 semiconductor Substances 0.000 description 2
- 241000282412 Homo Species 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000003796 beauty Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000035807 sensation Effects 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 230000001755 vocal effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L21/00—Speech or voice signal processing techniques to produce another audible or non-audible signal, e.g. visual or tactile, in order to modify its quality or its intelligibility
- G10L21/02—Speech enhancement, e.g. noise reduction or echo cancellation
- G10L21/038—Speech enhancement, e.g. noise reduction or echo cancellation using band spreading techniques
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L19/00—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis
- G10L19/02—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis using spectral analysis, e.g. transform vocoders or subband vocoders
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L21/00—Speech or voice signal processing techniques to produce another audible or non-audible signal, e.g. visual or tactile, in order to modify its quality or its intelligibility
- G10L21/02—Speech enhancement, e.g. noise reduction or echo cancellation
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/102—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or selection affected or controlled by the adaptive coding
- H04N19/103—Selection of coding mode or of prediction mode
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/102—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or selection affected or controlled by the adaptive coding
- H04N19/132—Sampling, masking or truncation of coding units, e.g. adaptive resampling, frame skipping, frame interpolation or high-frequency transform coefficient masking
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/134—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or criterion affecting or controlling the adaptive coding
- H04N19/136—Incoming video signal characteristics or properties
- H04N19/14—Coding unit complexity, e.g. amount of activity or edge presence estimation
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/169—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding
- H04N19/17—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object
- H04N19/176—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object the region being a block, e.g. a macroblock
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/169—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding
- H04N19/18—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being a set of transform coefficients
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/48—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using compressed domain processing techniques other than decoding, e.g. modification of transform coefficients, variable length coding [VLC] data or run-length data
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/50—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
- H04N19/59—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding involving spatial sub-sampling or interpolation, e.g. alteration of picture size or resolution
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/60—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using transform coding
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/85—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using pre-processing or post-processing specially adapted for video compression
- H04N19/89—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using pre-processing or post-processing specially adapted for video compression involving methods or arrangements for detection of transmission errors at the decoder
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/90—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using coding techniques not provided for in groups H04N19/10-H04N19/85, e.g. fractals
- H04N19/94—Vector quantisation
Definitions
- the present invention relates to a signal processing system, a signal processing device and method, a recording medium, and a program, and more particularly to a signal that accurately restores a high-frequency component of a narrow-band signal in which a high-frequency component is suppressed.
- the present invention relates to a processing system, a signal processing device and method, a recording medium, and a program. Background art
- Reference 1 Japanese Unexamined Patent Publication No. Hei 7-36490 (hereinafter referred to as Reference 1) discloses a method in which a lost harmonic component is generated by calculation based on an input audio signal spectrum. Addition methods are disclosed.
- Ref. 2 A paper by Saitama University at the Acoustical Society of Japan (hereinafter referred to as Ref. 2) announced a method of generating a voiced sound source waveform by autocorrelation calculation of an input speech signal and using this as a high-frequency component. Have been.
- Reference 3 discloses D / A at a higher sampling frequency than the input signal.
- a sample value interpolation method for performing the conversion is disclosed.
- Japanese Patent Application Laid-Open No. 6-118995 (hereinafter referred to as Document 4) associates a beta-quantization codepock of a wideband speech signal with a beta-quantization codepock of a narrow-band speech signal. Discloses a method for restoring a wideband audio signal. '
- Document 1 is based on the technique of removing the high-frequency signal lost in the compression process of a relatively wideband audio signal such as MP3 (Moving Picture Experts Group-1 Audio Layer-3). It is a predictive addition, and it is difficult to apply it to, for example, double the bandwidth of an extremely narrow band signal such as a telephone voice band.
- MP3 Motion Picture Experts Group-1 Audio Layer-3
- the technique of Reference 3 is an extension of the interpolation filter that sets the level of the interpolated sample value to a value that is considered appropriate when observing the change situation of several samples before and after when performing sample value interpolation processing to increase the sampling frequency. It is a technology on the wire and does not restore lost frequency components.
- the present invention has been made in view of such a situation, and it is an object of the present invention to accurately restore a high frequency component of a narrowband signal in which a high frequency component is suppressed.
- the first signal processing device includes: first signal input means for inputting a first signal having a wide frequency component; and a high frequency component of the first signal is suppressed.
- Second signal input means for inputting the obtained narrow-band second signal, extraction means for extracting high-frequency components from the first signal input by the first signal input means, A first determining means for determining a first class code based on the second signal input by the signal inputting means; and an extracting means for each first class code determined by the first determining means.
- Storage means for accumulating the extracted high frequency components, the second signal processing device comprising: third signal input means for inputting a narrow-band third signal in which the high frequency components are suppressed; Based on the third signal input by the third signal input means A second determining means for determining a second class code, corresponding to the second class code determined by the second determining means, the accumulation means A synthesizing means for synthesizing the accumulated high frequency component, a third signal input by the third signal input means, and an output means for outputting a synthesized signal generated by the synthesizing means. I do.
- the first signal processing device includes: a first signal input control step of controlling an input of a first signal having a wide frequency component; A second signal input control step for controlling the input of a narrow-band second signal in which the high-frequency component is suppressed; and a first signal input control step for controlling the input by the processing of the first signal input control step.
- the first signal processing device of the present invention comprises: first signal input means for inputting a first signal having a wide frequency component, and a narrow band in which a high frequency component of the first signal is suppressed. Second signal input means for inputting a high-frequency component from the first signal input by the first signal input means, and second signal input means for inputting a high-frequency component from the first signal input by the first signal input means. Determining means for determining a class code based on the input second signal; and storing means for storing the high frequency components extracted by the extracting means for each class code determined by the determining means. It is characterized by.
- a first signal input control step of controlling an input of a first signal having a wide frequency component, and a high frequency component of the first signal is suppressed
- a second signal input control step for controlling the input of a narrow-band second signal, and an extraction for extracting a high-frequency component from the first signal whose input is controlled by the processing of the first signal input control step A step for determining a class code based on the step and the second signal whose input is controlled by the processing of the second signal input control step; and for each class code determined by the processing of the determining step.
- an accumulating step of accumulating the high frequency components extracted by the processing of the extracting step is accumulating the high frequency components extracted by the processing of the extracting step.
- the program of the first recording medium comprises: a first signal input control step of controlling input of a first signal having a wide frequency component; and a high frequency component of the first signal is suppressed.
- a second signal input control step for controlling the input of the narrow-band second signal, and a high-frequency component from the first signal whose input is controlled by the processing of the first signal input control step.
- An extracting step of extracting, a class code is determined based on the second signal whose input is controlled by the processing of the second signal input control step, a determining step, and a class code determined by the processing of the determining step. And accumulating the high frequency components extracted by the processing of the extracting step.
- a first program includes a first signal input control step of controlling an input of a first signal having a wide frequency component, and a narrow signal in which a high frequency component of the first signal is suppressed.
- a second signal input control step of controlling input of a second signal in a band, and an extraction step of extracting a high frequency component from the first signal whose input is controlled by the processing of the first signal input control step.
- a determination step of determining a class code based on the second signal whose input is controlled by the processing of the second signal input control step, and an extraction for each class code determined by the processing of the determination step.
- a storing step of storing the high frequency components extracted by the processing of the step.
- An audio signal can be input to the input means of the first signal processing device of the present invention.
- the determining means can determine the class code based on the signal on which the weighting process based on the equal loudness characteristic is performed. .
- the determining means includes a class code based on the position of the average power value and the maximum power value of the voice signal spectrum. The decision can be made.
- the determining means includes: 1/2 of the lowest frequency of the high frequency component stored in the storage means, or 1 It is possible to determine the class code of a spectral component having a frequency equal to or higher than Z three times.
- An image signal can be input to the input means of the first signal processing device of the present invention.
- the determining means includes an AC average power value, a DC power value, and an AC component of a spectrum of the image signal.
- the class code can be determined based on the peak power value.
- the high frequency components extracted by the extraction means can be stored in the storage means of the first signal processing device of the present invention in association with the specific information.
- the specific information of the first signal processing device of the present invention may be a telephone number of a telephone.
- the signal processing apparatus may further include a generation unit configured to generate a second signal by thinning out the first signal at predetermined time intervals.
- a second signal processing device comprises: signal input means for inputting a narrow-band signal in which high-frequency components are suppressed; and determination means for determining a class code based on the signal input by the signal input means. And storage means for storing high frequency components, A high frequency component stored in the storage means corresponding to the class code determined by the stage; a synthesis means for synthesizing the signal input by the signal input means; and a synthesized signal generated by the synthesis means. Output means.
- a second signal processing method is based on a signal input control step of controlling input of a narrow-band signal in which a high-frequency component is suppressed, and a signal whose input is controlled by the signal input control step.
- a storage step of controlling the storage of the high frequency components and a storage control step corresponding to the class code determined by the processing of the determination step.
- the program of the second recording medium of the present invention is based on a signal input control step of controlling input of a narrow-band signal in which a high frequency component is suppressed, and a signal whose input is controlled by the processing of the signal input control step.
- the storage step is controlled by the storage control step corresponding to the class code determined by the determination step.
- a second program includes a signal input control step of controlling input of a narrow-band signal in which a high-frequency component is suppressed, and a signal whose input is controlled by the processing of the signal input control step.
- causing the computer to execute a synthesizing step of synthesizing the frequency component and the signal input by the processing of the signal input control step, and an output control step of controlling the output of the synthesized signal generated by the processing of the synthesizing step.
- An audio signal can be input to the input means of the second signal processing device of the present invention.
- the determining unit is configured to determine a class code based on the signal on which the weighting process based on the equal loudness characteristic is performed. be able to.
- the determining means includes a class code based on the position of the average power value and the maximum power value of the audio signal spectrum. The decision can be made.
- the determination means includes: 1/2 times the lowest frequency of the high frequency component stored in the storage means, or 1 It is possible to determine a class code of a spectrum component having a frequency three times or more.
- An image signal can be inputted to the input means of the second signal processing device of the present invention.
- the determining means includes an AC average power value, a DC power value, and an AC component peak power of a spectrum of the image signal.
- the class code can be determined based on the value.
- the high frequency components stored by the storage means of the second signal processing device of the present invention can be updated at a predetermined timing.
- the high frequency components can be stored in the storage means of the second signal processing device of the present invention in association with the specific information.
- the specific information of the second signal processing device of the present invention may be a telephone number of a telephone.
- the first signal processing device extracts a high-frequency component from the first signal having a wide frequency component, and outputs a high-frequency component of the first signal. Based on the narrow-band second signal whose frequency components have been suppressed, Four
- One class code is determined, and high frequency components are accumulated for each class code.
- the second class code is determined based on the narrow-band third signal in which the high-frequency component is suppressed, and the stored high-frequency band corresponding to the second class code is determined. The frequency component and the third signal are combined.
- a high-frequency component is extracted from the first signal having a wide frequency component, and the high-frequency component of the first signal is suppressed.
- a class code is determined based on the narrow band second signal thus obtained, and a high frequency component is accumulated for each class code.
- a class code is determined based on a narrow band signal in which a high frequency component is suppressed, and a high frequency component corresponding to the class code is determined.
- a narrow band signal is synthesized, and a synthesized signal is output.
- the signal processing device may be an independent device, or may be a block that performs signal processing of one device.
- FIG. 1 is a block diagram illustrating a configuration example of a voice broadband device according to a first embodiment of the present invention.
- FIG. 2 is a block diagram showing a configuration example of the class classification unit in FIG.
- FIG. 3 is a flowchart for explaining the audio band broadening process of the audio band broadening device of FIG.
- FIG. 4A is a diagram showing an example of data of the voice broadening process.
- FIG. 4B is a diagram showing an example of data of the voice broadening process.
- FIG. 4C is a diagram showing an example of data of the voice broadening process.
- FIG. 5A illustrates the basic principle of the classification process.
- FIG. 5B is a diagram illustrating the basic principle of the classification process. 2003/010824
- FIG. 6 is a flowchart illustrating the class classification processing in step S3 in FIG.
- FIG. 7 is a diagram illustrating an equal loudness curve of a sound.
- FIG. 8 is a diagram showing an A characteristic curve for correcting the frequency of the sound.
- FIG. 9 is a view for explaining the process of step S25 in FIG.
- FIG. 10 is a block diagram showing a configuration example of a speech signal learning device.
- FIG. 11 is a flowchart illustrating a learning process of the learning device in FIG. 10.
- FIG. 12A is a diagram illustrating an example of data of an audio signal in the learning process.
- FIG. 12B is a diagram illustrating an example of data of an audio signal in the learning processing.
- FIG. 12C is a diagram illustrating an example of data of an audio signal in a learning process.
- FIG. 12D is a diagram illustrating an example of audio signal data in the learning process.
- FIG. 12E is a diagram illustrating an example of data of an audio signal in a learning process.
- FIG. 12F is a diagram illustrating an example of data of an audio signal in a learning process.
- FIG. 13 is a flowchart illustrating the class classification processing in step S53 of FIG.
- FIG. 14 is a block diagram showing a configuration example of an image signal broadening apparatus according to the second embodiment of the present invention.
- FIG. 15 is a block diagram illustrating a configuration example of the class classification unit in FIG.
- FIG. 16A is a diagram showing an example of data of the image signal broadening process.
- FIG. 16B is a diagram showing an example of data of the image signal broadening process.
- FIG. 16C is a diagram showing an example of data of the image signal broadening process.
- FIG. 17 is a flowchart for explaining the class classification processing of the image signal broadening apparatus in FIG.
- FIG. 18 is a view for explaining the processing in step S92 of FIG.
- FIG. 19 is a block diagram showing a configuration example of the image signal learning device.
- FIG. 2OA is a diagram illustrating an example of data of an image signal in a learning process.
- FIG. 2 OB is a diagram showing an example of image signal data of the learning process.
- FIG. 2 OC is a diagram illustrating an example of data of an image signal in a learning process.
- FIG. 2OD is a diagram showing an example of data of an image signal in the learning process.
- FIG. 20E is a diagram illustrating an example of data of an image signal in a learning process.
- FIG. 20F is a diagram illustrating an example of data of an image signal in a learning process.
- FIG. 21 is a flowchart illustrating a class classification process of the learning device in FIG.
- FIG. 22 is a block diagram illustrating a configuration example of a transmission system according to the third embodiment of the present invention.
- FIG. 23 is a block diagram showing a configuration example of the mobile phone shown in FIG.
- FIG. 24 is a block diagram illustrating a configuration example of the transmission unit in FIG.
- FIG. 25 is a block diagram illustrating a configuration example of the receiving unit in FIG.
- FIG. 26 is a block diagram illustrating a configuration example of the encoding unit in FIG. 24 when the learned table is provided as fixed information.
- FIG. 27 is a block diagram illustrating a configuration example of the decoding unit in FIG. 25 when the learned table is given as fixed information.
- FIG. 28 is a block diagram illustrating a configuration example of the encoding unit in FIG. 24 when the learned table is updated at a predetermined timing.
- FIG. 29 is a block diagram illustrating a configuration example of the decoding unit in FIG. 25 when the learned table is updated at a predetermined timing.
- FIG. 30 is a flowchart illustrating a transmission process of the transmission unit in FIG.
- FIG. 31 is a flowchart for explaining the memory data transmission process in step S202 of FIG.
- FIG. 32 is a flowchart illustrating the encoding process in step S205 of FIG.
- FIG. 33 is a flowchart illustrating a reception process of the reception unit in FIG. 25.
- FIG. 34 is a flowchart for explaining the memory data update process in step S252 of FIG.
- FIG. 35 is a flowchart for explaining the decoding process in step S254 of FIG.
- FIG. 36 is a block diagram showing a configuration example of a computer according to an embodiment of the present invention.
- FIG. 1 shows a configuration example of a voice broadband device according to a first embodiment of the present invention.
- a narrow-band audio signal having a low-frequency component (no high-frequency component) is input to the frame cutout unit 11.
- the input audio signal is divided into predetermined frames (unit time) and output to a DCT (Discrete Cosine Transform) converter 12.
- DCT Discrete Cosine Transform
- the DCT converter 12 converts the audio signal on the time axis input from the frame cutout unit 11 into a signal on the frequency axis for each frame, and the classification unit 13 and the spectrum coupling unit 1 Output to 6 respectively.
- M-DCT Modified DCT
- M-DCT Modified DCT
- DFT Discrete Fourier transform
- the class classification unit 13 calculates a class code by class classification processing (the details of which will be described later with reference to FIG. 6) based on the input spectrum signal, and outputs the class code to the learned table 14. .
- the learned table 14 is a table memory having class codes as addresses. Each class code address is obtained by a learning device described later with reference to FIG. The vector data of the obtained audio signal spectrum of the high frequency component is stored. The learned table 14 reads the vector data of the audio signal spectrum of the high-frequency component corresponding to the class code address input from the class classification unit 13 and outputs the read data to the gain adjustment unit 15.
- the gain adjusting unit 15 adjusts the gain of the audio signal spectrum data of the high frequency component input from the learned table 14 and the low frequency component combined by the spectrum combining unit 16. Adjust to match the gain of the audio signal spectrum data.
- the spectrum coupling unit 16 converts the beta data of the audio signal spectrum of the high frequency component input from the gain adjustment unit 15 into the low frequency component of the low frequency component input from the DCT conversion unit 12. It is combined with the last row of the beta data of the audio signal spectrum and output to the inverse DCT converter 17.
- the inverse DCT transform unit 17 converts the spectrum data on the frequency axis into the spectrum on the time axis by performing the inverse transform of the M-DCT transform, and outputs the spectrum data to the frame combining unit 18. .
- the frame combining unit 18 outputs an audio signal after performing a process of adding a frame overlap portion to prevent audio from becoming discontinuous at a frame boundary.
- FIG. 2 shows a detailed configuration example of the class classification unit 13 in FIG.
- the DCT spectrum data output from the DCT converter 12 and subjected to the M-DCT conversion is input to the power value converter 31.
- the input DCT spectrum data is converted into a power value, further converted into a power dB (decibel) value, and output to the weighting processing unit 32.
- the weighting processing unit 32 performs a weighting process on the input DCT spectrum data in consideration of the equal loudness characteristic described later with reference to FIG. 7, and outputs the data to the spectrum extraction unit 33.
- the spectrum extraction unit 3 3 converts the DCT spectrum data of the input low frequency component into a DCT that is at least 1/2 or 1/3 times the lowest frequency of the high frequency component to be restored.
- the spectrum data is extracted and output to the sub-band division unit 34.
- the sub-band division unit 34 divides the DCT spectrum data extracted by the spectrum extraction unit 33 into a predetermined number of sub-bands and outputs the sub-band to the quantization unit 35.
- the quantization unit 35 combines the DCT spectrum data divided into sub-bands in sub-band units, and obtains an average value of the input power values (extracted by the spectrum extraction unit 33). Then, the average value is quantized to generate a class code, the subband position that takes the maximum power value is determined, and the class code is generated based on the position.
- the two class codes are output to the class code determination unit 36.
- the class code determination unit 36 multiplexes the class code based on the average power value input from the quantization unit 35 and the class code based on the maximum power position, generates one class code, and has learned Output to Table 14 ( Figure 1).
- step S1 the frame cutout unit 11 frames the audio signal, which has been input to the audio broadband device 1 and in which the high frequency components are suppressed, into frames. That is, the audio signal is divided into blocks every predetermined frame (unit time).
- step S2 the DCT conversion section 12 performs M-DCT conversion on the audio signal framed in step S1.
- FIG. 4A shows data obtained by subjecting a one-frame audio signal to M-DCT conversion (hereinafter, referred to as DCT spectrum data).
- step S3 the class classification unit 13 performs a class classification process described later with reference to FIG. 6, and calculates a class code.
- step S4 the learned table 14 reads out the DCT vector data of the high frequency component spectrum stored in the address corresponding to the class code output from the classifying unit 13 in step S3. .
- FIG. 4B shows the DCT spectrum data of the high frequency component read in this way.
- step S5 the spectrum coupling unit 16 converts the DCT spectrum data of the low-frequency component (the portion shown as a spectrum in FIG. 4A) input to the speech broadband device 1, DCT of high frequency component read from learned table 14 Combine the spectral data (shown as spectrum ⁇ in Figure 4B). That is, as shown in FIG. 4C, the front end (low-frequency side) of the spectrum is connected to the rear end (high-frequency side) of the spectrum A L. Since the number of data in one frame output by the DCT converter 12 is N and the number of data read from the learned table 14 is also N, the number of combined data is 2 N.
- step S6 the inverse DCT transform section 17 performs inverse M-DCT transform on the spectrum-coupled DCT spectrum data.
- step S7 the frame combiner 18 performs overlap addition processing on the frame boundary, outputs the result, and ends the audio band broadening processing. This overlapping process prevents the sound from becoming discontinuous even at the frame boundaries.
- Figure 5A shows the band width 2B.
- Fig. 5B shows the bandwidth B generated by removing the harmonic components from the teacher DCT spectrum. This is a student DCT spectrum having only a narrow band low frequency component.
- Classification processing is performed based on the assumption that the amount of high-frequency components contained in the narrow-band signal and the frequency components whose spectral shape is to be restored have a link relationship.
- the band width B in FIG. 5B is assumed.
- a signal equal to three times the harmonic component is included in the region Q of approximately half of the high frequency side of the band width 2 B 0 of FIG. 5 A of It is expected that.
- the spectrum hardly exists in the region P, it is expected that the signal is not likely to exist in the region Q.
- step S22 the weighting processing unit 32 performs a weighting process on the decibel value PWR dB of the power calculated in step S23 in consideration of the characteristics of the human ear.
- Fig. 7 The relationship between the loudness of sound (the amount of sensation) and the sound pressure level (physical amount) that can be subjectively felt by humans can be represented by an equal loudness curve as shown in Fig. 7. This shows the relationship between the sound pressure level and the frequency of a pure sound that a person with normal hearing perceives as equal. As this curve shows, our human ear is less sensitive at lower and higher frequencies.
- step S2 The process of performing frequency correction based on the characteristics of the human ear is described in step S2.
- step S23 the spectrum extraction unit 33 generates a high-frequency component to be restored from the input low-frequency component spectrum data.
- the spectrum data having a frequency equal to or more than 1/2 or 1/3 of the lowest frequency is extracted and output to the subband division unit 34. This is performed based on the basic principle of the class classification processing described above.
- the lowest frequency in the region Q in FIG. 5A as the high frequency component to be restored is f.
- Scan Bae-vector is extracted in the half of the frequency f Q / 2 or more regions P.
- subband splitting unit 3 4 scan Bae spectrum region P from the frequency f Q / 2 of the scan Bae-vector data (Fig. 5 B extracted in step S 2 3 until the frequency f Q Is divided into a predetermined number of subbands, for example, 32 subbands, and the spectrum data is collected in subband units.
- step S25 as shown in FIG. 9, the average power value P AV of the 32 entire sub-bands and the sub-band with the largest decibel power value PWR dB of the 32 sub-bands
- the position B n is determined.
- the range of the decibel power value of the spectrum is, for example, 0 to 127 [dB].
- step S 2 6 the class code determining unit 3-6, and the class code based on the average power value P AV output in step S 2 5, multiplexes a class code by the maximum power position B n.
- One of the four types is output as a class code.
- the learning device 50 is provided, for example, by a manufacturer that manufactures the voice broadband device 1.
- a wide-band audio signal including a high-frequency component serving as teacher data and a narrow-band audio signal in which the high-frequency component of the teacher data is suppressed are prepared as student data.
- the student data is generated by thinning out the sample of the low frequency component generated by removing the high frequency component of the teacher data by half by the decimation circuit 65. Therefore, assuming that the frame length of the student data input to the frame cutout unit 66 is N, the frame length of the teacher data input to the frame cutout unit 61 is 2N. At this time, the data of each frame is synchronized.
- the functions of the frame cutout unit 61 and the frame cutout unit 66, and the DCT conversion unit 62 and the DCT conversion unit 67 are the same as those of the frame cutout unit 11 and the DCT conversion unit 12 in FIG. .
- the high frequency component extraction section 63 extracts half of the spectrum of the teacher data on the high frequency side, and outputs it to the addition section 64.
- the class classifying unit 68 generates the same class code as described above in the class classifying unit 13 in FIG. 1, further calculates the class appearance frequency, and outputs it to the integration table 69.
- the integration table 69 inputs the class code output by the class classification unit 68 as an address, reads out the DCT vector data stored at that address, and outputs it to the addition unit 64.
- the adder 64 adds the DCT vector data of the N high-frequency components output from the high-frequency component extractor 63 and the N DCT vector data output from the integration table 69 by vector addition. Then, the result is output to the same class code address as when the integration table 69 was output. Since one class code address stores N DCT vector data, the accumulated table 69 becomes a table having the number of data of (N X class number).
- the integration table 69 divides the vector data stored in each class code address by the class appearance frequency output from the class classification unit 68 to obtain the learned table 70. Output to
- the learned table 70 stores (the number of NX classes) DCT vector data output from the integration table 69.
- the data in the learned table 70 is transferred to the voice broadband device 1 via the network, and stored in the learned table 14.
- the data of the learned table 70 is written into a semiconductor memory or the like, and is incorporated as the learned table 14 of the audio band widening apparatus 1. Since the configuration of the classifying unit 68 of the learning device 50 is the same as that of the classifying unit 13 of the speech broadband device 1, the configuration of the classifying unit 13 in FIG.
- the configuration of the 0 class classification unit 68 is also referred to as appropriate.
- step S51 the teacher data spectrum and the student data spectrum are framed. That is, the frame cutout unit 61 frames the input teacher data for each unit time, and performs 2N samplings for one frame.
- the thinning circuit 65 generates student data by extracting only low frequency components from the teacher data and thinning out every other low frequency component.
- the frame cutout unit 66 frames the student data for each unit time, and performs N samplings for one frame.
- step S52 the DCT converter 62 and the DCT converter 67 perform M-DCT conversion on the teacher data and the student data, respectively.
- Figure 12A shows the DCT spectrum data after the teacher data has been M-DCT transformed
- Figure 12B shows the DCT spectrum data after the student data has been M-DCT transformed. I have.
- the DCT spectrum of the student data in Fig. 12B corresponds to the low frequency component of the M-DCT spectrum of the teacher data in Fig. 12A.
- step S53 the class classification unit 68 performs a class classification process described later with reference to FIG. 13 based on the spectrum of the student data, and calculates the class code and the appearance frequency of the class. Output to integration table 69.
- step S54 the integration table 69 reads the vector data stored in the address corresponding to the class code output by the class classification unit 68.
- step S55 the high frequency component extraction unit 63 extracts the high frequency component of the teacher data input from the DCT conversion unit 62.
- Figure 12C shows the extracted DCT spectrum of the high frequency component. This corresponds to the high frequency side portion R in the entire DCT spectrum of the teacher data in Fig. 12A.
- step S56 the addition unit 64 extracts the vector data (data shown in FIG. 12D) read from the integration table 69 in step S54 and the teacher data in step S55. The vector data of the obtained high-frequency component vector data (data shown in Fig. 12C) is added.
- FIG. 12E shows the DCT spectrum data after the vector addition in step S56.
- step S57 the adding section 64 adds the vector data calculated in step S56 (the data shown in FIG. 12E) to the address of the same class code as when the integration table 69 was read. Store in place.
- step S58 it is determined whether the learning process has been completed. If it is determined that the learning process has not been completed, the process returns to step S51, and the subsequent processes are repeated. If it is determined that the learning process has been completed, the process proceeds to step S59.
- step S59 the integration table 69 stores the N vector data at the address of each class code in the classifying unit 6
- the average value of the vector data is calculated by dividing by the class appearance frequency of the corresponding class code output from 8 and output to the learned table 70, stored, and the learning process is terminated.
- FIG. 12F shows an example of N pieces of solid data stored in the address of one class code in the learned table 70.
- step S53 of FIG. 11 will be described with reference to the flowchart of FIG.
- steps S71 to S76 is the same as that of steps S21 to S26 of the class classification processing in the speech band broadening apparatus of FIG.
- a class code based on the average power value P AV of the student data and a class code based on one of the maximum powers B n are multiplexed.
- step S77 the class code determination unit 36 (similar to that of the voice broadband device 1 in FIG. 2) counts the class code determined in step S76 by one. The number of occurrences of each class code is output to the integration table 69, and the classification process is completed.
- the class classification process of the voice broadening device 1 and the class classification process of the learning device 50 are the same until the class code is calculated, and thereafter, only whether or not to calculate the class appearance frequency is determined. different.
- FIG. 14 shows an example of the configuration of an image signal broadening device according to a second embodiment of the present invention.
- the above-described audio signal broadening apparatus 1 has an input signal which is a one-dimensional audio signal and performs processing in units of frames, whereas the image signal broadening apparatus 80 has a two-dimensional image signal. Therefore, processing is performed in units of blocks (for example, 8 ⁇ 8 pixels).
- the block cutout 91, DCT transform 92, class classifier 93, trained table 94, combining 95, and inverse DCT transform 96 in Fig. 14 are the frame cutout 1 in Fig. 1. 1, which has basically the same functions as the DCT transform section 12, class classification section 13, trained table 14, spectrum coupling section 16, and inverse DCT transform section 17. Therefore, the description is omitted.
- the gain adjuster 15 and the frame combiner 18 shown in FIG. 1 are omitted in the case of an image signal because they are not necessarily required.
- FIG. 15 shows a detailed configuration example of the class classification unit 93 of FIG.
- the DCT data subjected to M-DCT conversion from the DCT converter 92 is input to the power value converter 101.
- the input DCT data is converted into a power value, further converted into a decibel value of the power, and output to the quantization unit 102 in the same manner as in the above-described audio broadband device 1.
- the quantization unit 102 calculates the DC component power value, the average power value of the AC component, and the position of the DCT data that takes the peak value of the AC component for the DCT data in the DCT block, and the class code based on them is calculated. Is generated and output to the class code determination unit 103.
- the class code determination unit 103 multiplexes the class code based on the input quantized data of the DCT block to form one class code, and outputs the class code to the learned table 94 (FIG. 14) for storage. .
- step S1 the block cutout unit 91 cuts out the image signal, in which the high frequency components are suppressed, input to the image signal broadening device 80, in units of blocks (for example, 22 pixels).
- step S2 the 0 ⁇ conversion section 92 performs M-DCT conversion on the image signal blocked in step S1.
- FIG. 16A shows data obtained by subjecting an image signal of one block to M-DCT conversion (hereinafter, referred to as DCT data).
- step S3 the class classification unit 93 performs a class classification process described later with reference to FIG. 17 to calculate a class code.
- step S4 the learned table 94 reads the DCT coefficient vector data stored in the address corresponding to the class code output from the classifying unit 93 in step S3.
- FIG. 16B shows the vector data of the DCT coefficients of the high frequency components read at this time.
- step S5 the combining unit 95 generates the DCT data of the low frequency component (FIG. 16A) input to the image signal broadening device 80 and the high frequency component read from the learned table 94.
- the DCT data of the component (Fig. 16B) is added by vector.
- Figure 16C shows the combined DCT block data.
- step S6 the inverse 1) ( ⁇ transform section 96 performs inverse M-DCT transform on the combined DCT data.
- step S7 the overlap processing in step S7 does not necessarily need to be performed, and thus the processing in step S7 is not performed and the image signal broadening processing ends.
- the class classification processing performed by the class classification unit 93 will be described with reference to the flowchart in FIG.
- step S91 the power value converter 101 converts the DCT data input from the DCT converter 92 into a power value, and further converts it into a decibel value of power. This process is the same as the above-described case of the audio signal.
- the quantization unit 102 calculates the AC average power value and the DC power value of the DCT block data, and detects the AC component peak power position.
- a DCT block BL of 4 ⁇ 4 pixels as shown in FIG. 18 is input, the DCT data of each pixel is c [n], and the power range is 0 to 127. [dB].
- the AC average power value is the average value of the DCT data c [l] to c [15] in the area AC in Fig. 18, and the DC power value is the DCT data c [0] in the area DC in Fig. 18.
- Is the value of The AC component peak power position is n (one of 1 to 15) having the maximum power value among the DCT data c [l] to c [15].
- step S93 the class classification unit 93 multiplexes the class code based on the three values obtained in step S92 as one class code, outputs the multiplexed class code to the learned table 94, and performs the class classification process.
- FIG. 19 is a configuration example of a learning device 120 that generates the high frequency component beta data stored for each class code, which is written in the learned table 94 of the image signal broadening device 80.
- Block cutout section 131, DCT conversion section 1332, high frequency component extraction section 1333, addition section 134, decimation circuit 1 35, block cutout section 1 36, DCT conversion section 1 in Fig. 19 3 7, Class classification section 1 3 8 and integrated tape 1 3 9 are the frame cutout section 6 1, DCT conversion section 6 2, high frequency extraction section 6 3, addition section 6 4, Circuit 65, Frame cutout 66, DCT converter 67, the classifying unit 68, and the accumulation table 69 have basically the same configuration and function, and a description thereof will be omitted.
- step S51 DCT data of teacher data and DCT data of student data are cut out into blocks. That is, the block cutout unit 1331 blocks the input teacher data for every 2N ⁇ 2N pixels.
- the thinning circuit 135 generates student data by extracting only low-frequency components from the teacher data and thinning out the low-frequency components alternately in the horizontal and vertical directions. .
- the block cutout unit 1336 blocks the student data for every N ⁇ N pixels.
- step S52 the DCT converters 13 and 13 perform M-DCT conversion on teacher data and student data, respectively.
- Figure 2 OA shows the DCT data (4 x 4 data) after the teacher data was converted to M-DCT
- Figure 20B shows the DCT data (M-DCT converted from the student data). 2 X 2 data).
- step S53 the classifying unit 138 performs the classifying process described below with reference to FIG. 21 on the DCT data as the student data, calculates the class code and the appearance frequency of the class, and Output to Table 1 39.
- step S54 the integration table 1339 reads the vector data stored in the address corresponding to the class code output by the classification unit 1338.
- step S55 the high frequency component extraction section 133 extracts the high frequency component of the teacher data.
- FIG. 20C shows the extracted high-frequency components. This is the part P (4 ⁇ 4 pixels excluding the upper left 2 ⁇ 2 pixel of the teacher data of FIG. 20A). ).
- step S56 the adder 1 34 outputs the integration table 1 in step S564.
- FIG. 20E shows the DCT data after the vector addition in step S56.
- step S57 the adder 1334 converts the vector data calculated in step S56 (the data shown in FIG. 20E) into the address of the same class code as when the integration table 1339 was read. In the location.
- step S58 it is determined whether the learning process has been completed. If it is determined that the learning process has not been completed, the process returns to step S51, and the subsequent processes are repeated. If it is determined that the learning process has been completed, the process proceeds to step s
- step S59 the integration table 1339 stores (3 NXN) vector data at the address of each class code in the class The average value of the vector data is calculated by dividing by the class appearance frequency of the corresponding class code output from the classifying unit 1338, output to the learned table 140, and the process ends.
- Figure 2 OF is the learned table 1
- steps S101 to S103 is the same as that of steps S91 to S93 of the class classification processing in the image signal broadening apparatus of FIG. 17, and a description thereof will be omitted. That is, the process of calculating the class code is performed in the same manner as in the image signal broadband device. By this processing, the class code based on the AC average power value, the DC power value, and the AC component peak power position is multiplexed as one class code.
- the class code determination unit 103 increases the count of the class code determined in step S103 by one, and counts the number of occurrences of the class code and the integration table 13 Output to 9 and end the classification process.
- FIG. 22 shows a transmission system according to the third embodiment to which the present invention is applied (a system refers to a device in which a plurality of devices are logically aggregated, and whether or not the devices of each configuration are in the same housing. Does not matter).
- the portable telephone 30 and 3 0 1 2 performs transmission and reception by radio, each base station 3 0 2t and 3 0 2 2 , by performing the transmission and reception to and from the switching station 3 0 3, finally, between the mobile telephone 3 0 3 0 1 2, the base station 3 0 2 and 3 0 2 2, the switching center to the parallel beauty Voice transmission and reception can be performed via 303.
- the base station 3 0 2i and 3 0 2 2 may be the same base station, or may be a different base station.
- FIG. 23 shows a configuration example of the mobile phone 30 of FIG. Since the portable phone 3 0 1 2 are also configured similarly to the cellular phone 3 0 1 ⁇ described below, and a description thereof will be omitted.
- Antenna 3 1 1 receives the radio waves from the base station 3 0 2L or 3 0 2 2, the reception signal, and supplies the modem unit 3 1 2, a signal from the modem part 3 1 2, Telecommunications in, and transmits to the base station 3 0 2i or 3 0 2 2.
- the modulation and demodulation unit 312 demodulates a signal from the antenna 311 by, for example, a CDMA (Code Division Multiple Access) method or the like, and supplies a demodulated signal obtained as a result to the reception unit 314. Further, the modulation and demodulation unit 312 modulates the transmission data supplied from the transmission unit 3113 by, for example, a CDMA method or the like, and supplies a modulated signal obtained as a result to the antenna 311.
- CDMA Code Division Multiple Access
- the transmission unit 3 13 is configured to encode a user's voice input thereto. Processing is performed, transmission data is obtained, and supplied to the modulation and demodulation unit 312.
- the receiving section 314 receives the received data, which is a demodulated signal from the modem section 314, and decodes and outputs high-quality sound.
- the operation unit 315 is operated by the user when inputting a telephone number of a call destination, a predetermined command, or the like, and an operation signal corresponding to the operation is transmitted by the transmission unit 313 or the reception unit 314. Supplied to
- FIG. 24 shows an example of the configuration of the transmission section 3 13 in FIG.
- the microphone 3 2 1 receives the user's voice, and the microphone 3 2 1 converts the user's voice into an A / D
- the / 0 converter 322 converts the analog audio signal from the microphone 321 into AZD, thereby converting the analog audio signal into digital audio data and outputting the digital audio data to the encoder 323.
- the encoding unit 3 22 3 encodes the audio data from the A / D conversion unit 3 22 by a predetermined encoding method, multiplexes the class code, and controls the transmission of the encoded audio data obtained as a result. Output to section 3 2 4.
- the management unit 327 manages the telephone number of the call destination, the telephone number of the caller, and other necessary information that are input by operating the operation unit 315 as necessary. , And outputs the result to the encoding unit 323 and the transmission control unit 324.
- the transmission control unit 324 controls transmission of the encoded voice data output from the encoding unit 323 and the data output from the management unit 327. That is, the transmission control unit 3224 selects the encoded voice data output from the encoding unit 3223 or the data output from the management unit 3227, and, at a predetermined transmission timing, transmits the data as a modulation / demodulation unit as transmission data. Output to 3 1 2 (Fig. 23).
- FIG. 25 shows an example of the configuration of the receiving section 314 of FIG.
- Received data as a demodulated signal output from the modulation and demodulation unit 312 in FIG. 23 is supplied to the reception control unit 331, and the reception control unit 331 receives the received data. Then, when the received data is coded voice data, the reception control unit 331 supplies the coded voice data to the decoding unit 3332. Alternatively, when the received data is the telephone number or other information of the caller, the reception control unit 331 transmits such information to the management unit 335 or the transmission unit as necessary. 3 1 3 (to the management section 3 2 7).
- the decoding unit 3332 separates the encoded audio data supplied from the reception control unit 331 from the class code, adds a high frequency component, and converts the resulting decoded audio data into a DZA ( (Digital / Analog) converter 3 3 3
- the D / A conversion section 333 performs D / A conversion on the digital decoded audio data output from the decoding section 332, and supplies an analog audio signal obtained as a result to the speaker 334.
- the speaker 334 outputs a sound corresponding to the sound signal from the DZA converter 333.
- the management unit 335 receives the telephone number of the caller from the reception control unit 331 when receiving a call, and supplies the telephone number to the decoding unit 332.
- FIG. 26 and FIG. 27 show configuration examples of the encoding unit 323 and the decoding unit 332 when the learned table is given as fixed information.
- the functions of the frame cutout unit 351 and the DCT conversion unit 352 are the same as those of the frame cutout unit 11 and the DCT conversion unit 12 of the voice broadband device 1 in FIG.
- the high-frequency component removal unit 355 3 removes the high-frequency components of the input DCT spectrum data and generates the low-frequency components to class code multiplexing with the class classification unit 354. Output to part 3 5 5
- the compression rate depends on the removal rate of the high frequency components. For example, if the band is compressed in half, the data amount is reduced to half. The final data amount is obtained by adding several codes of the class codes multiplexed in the Las code multiplexing unit 355.
- the classifying unit 354 applies the DCT spectrum data of the low-frequency component input from the high-frequency component removing unit 353 to the classifying unit 13 of the audio band widening apparatus 1. The same class classification processing as described above is performed, and the class code is output to the class code multiplexing unit 355.
- the class code multiplexing unit 355 multiplexes the class code output from the classifying unit 354 and the DCT spectrum data of the low frequency component output from the high frequency component removing unit 353. And outputs it to the inverse DCT converter 356.
- the class code multiplexing unit 355 multiplexes the class code output from the classifying unit 354 and the DCT spectrum data of the low frequency component output from the high frequency component removing unit 353. And outputs it to the inverse DCT converter 356.
- By multiplexing the class code with DCT spectrum data of low-frequency components it is possible to improve the error resistance of the class code on the code transmission line, and to improve the decoding capability of the decoding unit 33 (Fig. 27) described later. It is also possible to detect and correct errors in the multiplexed class code by examining the degree of coincidence between the multiplexed class code and the class code obtained by performing the class classification process again.
- ⁇ Transformer 356 performs inverse M-DCT transform on the multiplexed data of the input class code and DCT spectrum data of low frequency components, and outputs it to frame combiner 357 I do.
- the frame combining section 357 performs the same frame overlap processing as in the frame combining section 18 of the voice broadband device 1, and transmits the coded voice data to the transmission control section 32 4 (FIG. 24). Output.
- the frame cut-out unit 37 1 and the DCT conversion unit 37 2 have the same functions as the frame cut-out unit 11 and the DCT conversion unit 12 of the voice broadband device 1 in FIG. Omitted.
- the class code separation unit 373 separates the multiplexed DCT spectrum data into a class code and DCT spectrum data of low frequency components, and generates low frequency components.
- the DCT spectrum data of the corresponding minutes is output to the spectrum coupling section 376, and the class code is output to the learned table 374.
- the learned table 374 reads the DCT spectrum data of the high frequency components stored in the address corresponding to the class code output by the class code separation unit 373, and outputs it to the gain adjustment unit 375 I do.
- the gain adjuster 3775 like the gain adjuster 15 of the audio broadband device 1, performs learning so as to match the gain of the low-frequency component spectrum combined by the spectrum combiner 3776.
- the gain adjustment is performed on the spectrum of the high frequency component output from the processing table 374, and the spectrum is output to the spectrum coupling unit 376.
- the spectrum combining unit 3756 includes the DCT spectrum data of the low-frequency component output from the class code separating unit 3773 and the DCT spectrum of the high-frequency component output from the gain adjusting unit 3775.
- the vector data is combined and output to the inverse DCT converter 377.
- Inverse DCT transform section 377 performs inverse M-DCT transform of the input spectrum data, and outputs the result to frame combining section 378.
- the frame combining section 378 performs frame overlap processing and outputs the result to the D / A conversion section 333 as a wideband audio signal.
- Fig. 28 and Fig. 29 the learned table that stores the spectrum data of the high frequency components is stored in association with each telephone number of the caller, and the table information for the caller is updated each time it is used
- the same reference numerals are given to portions corresponding to those in FIGS. 26 and 27, and the description thereof will be omitted below.
- the spectrum dividing section 3881 divides the input DCT spectrum data into a high frequency component and a low frequency component, and outputs the high frequency component to the table memory 382. Then, the low frequency components are output to the class classification section 354 and the class code multiplexing section 355.
- the class classification unit 354 calculates the class code and outputs the class code to the table memory 382 and the class code multiplexing unit 355.
- the table memory 382 stores the DCT spectrum data of the high frequency component input from the spectrum dividing section 381 in an address corresponding to the class code output from the class categorizing section 354.
- the contents of the table memory are transferred to a table memory 391 (FIG. 29) of the decoding unit 3332 described later at a predetermined timing.
- the spectrum data of the high frequency component for each caller is stored in the table memory 38 2. It becomes. Since the contents of the table memory 382 are a kind of statistical information, it is expected that the more information is entered, the more likely it will be to converge to an appropriate bias in class code units. As the number increases, higher-precision decoding will be performed.
- Other configurations in FIG. 28 are the same as those in FIG.
- the table memory 391 reads out DCT spectrum data of a high frequency component stored at a position designated as an address corresponding to the class code output from the class code separation section 3733, and a gain adjustment section. Output to 3 7 5
- the mobile telephone 3 0 1 lambda transmitting unit 3 1 3 of the transmission process will be described which performs a source.
- Step S 2 0 the user operates the operating unit 3 1 5 (2 3), entering the mobile phone 3 0 1 2 phone number as the called party, the management unit 3 2 7, call Detect the start of.
- the transmission unit 313 performs a memory data transmission process described later with reference to FIG. That is, the transmission unit 3 1 3, table memory 3 8 2 (2 8) a DCT vector data of a high frequency component spectrum stored in receiver side of the mobile phone 3 0 1 2 table memory 3 9 Send to 1 (Fig. 29) and switch to voice call.
- step S203 the microphone 3221 inputs the user's voice.
- step S204 the A / D conversion section 322 performs A / D conversion on the audio signal input from the microphone 321 and outputs it to the encoding section 323 as a digital audio signal.
- step S205 the encoding unit 3233 performs an encoding process. The details will be described later with reference to FIG. 32.
- the audio signal input from the A / D converter 322 is divided into a high frequency component and a low frequency component. Then, the high frequency components are stored in the table memory 382, and the low frequency components are multiplexed with the class code, and output to the transmission control unit 324 as encoded voice data.
- the transmission control section 324 transmits the encoded voice data input from the encoding section 323 to the modem section 313 (FIG. 23).
- step S207 the management unit 327 determines whether or not the call has ended. If it is determined that the call has not been completed, the management unit 327 returns the process to step S203, and the subsequent processes are repeated. When it is determined that the call has ended, the management unit 327 ends the transmission process.
- step S 2 2 transmission control unit 3 2 4, the cellular phone 3 0 1 2 telephone numbers entered by the operation unit 3 1 5 is operated, by outputting as transmission data, the mobile phone A call to 3 0 1 2 is made.
- step S 2 2 transmission control unit 3 2 4 establishes a communication link between the terminating side of the mobile phone 3 0 1 2, the process proceeds to step S 2 2 3.
- step S223 the management unit 327 reads out the data stored in the table memory 382 of the encoding unit 323 and supplies it to the transmission control unit 324. Further, in step S223, the transmission control section 324 selects the memory data from the management section 327 and transmits it as transmission data. The memory data is transmitted together with update information indicating the date and time when the memory data was obtained by learning.
- step S 2 2 3 to S 2 2 4 the management section 3 2 7 determines ready notification is whether transmitted from the cellular phone 3 0 1 2 at the incoming side.
- step S 2 2 4 such Ready notification, whether sent is determined from the mobile phone 3 0 1 2. If it is determined in step S224 that the preparation completion notification has not been transmitted, the process returns to step S224 and waits until the preparation completion notification has been transmitted. Then, in step S224, when it is determined that the preparation completion notification has been transmitted, the process proceeds to step S225, where the transmission control unit 324 selects the output of the encoding unit 323. As a result, voice communication is enabled, and the memory data transmission process ends.
- step S205 in FIG. 30 will be described with reference to the flowchart in FIG.
- step S231 the frame cutout unit 351 converts the audio signal input from the A / D conversion unit 322 into a frame.
- the audio signal input here has both high frequency components and low frequency components.
- step S232 the DCT conversion section 352 performs M-DCT conversion on the audio signal framed in step S231 and converts the signal into DCT spectrum data. 8 Output to 1.
- step S233 the spectrum dividing section 3811 divides the DCT spectrum data into a high frequency component and a low frequency component, and divides the high frequency component into table memory. And the low-frequency components are output to the classifying section 354.
- step S 234 the class classification section 354 determines the class code by the class classification processing, and outputs it to the class code multiplexing section 355 and the table memory 382.
- the class classification process here is the same as the class classification process of the audio broadband device 1 described above with reference to FIG. 6, and a description thereof will be omitted.
- step S 235 the table memory 382 stores the DCT of the high frequency component input from the spectrum division section 381 into the address corresponding to the class code output from the classification section 354. Stores spectrum data.
- step S 236 the class code multiplexing section 355 converts the class code input from the class classification section 354 and the DCT spectrum of the low frequency component input from the spectrum division section 381.
- the vector data is multiplexed and output to the inverse DCT converter 356.
- step S 237 the inverse 0 ( ⁇ transform section 356 performs inverse DCT transform on the multiplexed DCT spectrum data of the low frequency components.
- step S 2308 the frame combining section 357 performs an overlap process to create encoded voice data.
- step S251 the management unit 335 detects the reception of the radio wave via the reception control unit 331.
- step S252 the receiving unit 314 performs a memory data update process described later with reference to FIG. That is, the receiving unit 314 updates the DCT vector data of the high frequency component spectrum stored in the table memory 391 (FIG. 29), and performs voice communication (encoded voice data). Switch to.
- step S 253 when the reception data input to reception control section 331 is switched to encoded audio data, reception control section 331 converts the encoded audio data to decoding section 33. Output to 2.
- step S254 the decoding unit 332 performs a decoding process. As will be described later in detail with reference to FIG. 35, by this processing, the decoding unit 332 separates the coded audio data in which the class code is multiplexed into a class code and a low frequency component. The high frequency components stored at the address corresponding to the class code are read out, combined with the low frequency components, and output as a wideband audio signal.
- step S255 the D / A conversion section 333 performs D / A conversion on the wideband audio signal input from the decoding section 332, and outputs the signal to the speaker 334.
- step S256 the speaker 3334 outputs the analog audio signal input from the D / A conversion section 3333.
- step S257 the management unit 335 determines whether or not the call has ended. When it is determined that the call has not ended, the management unit 335 returns the process to step S253, and the subsequent processes are repeated. When it is determined that the call has ended, the management unit 335 ends the reception processing.
- step S271 the reception control unit 3331 determines whether or not the user has set the operation unit 315 to the off-hook state by operating the operation unit 315. Return to step S27.
- step S 272 the reception control unit 331 establishes a communication link, and proceeds to step S 273.
- step S 273 the reception control section 331 receives the reception data including the latest memory data transmitted from the transmitting-side mobile phone 30, and supplies the reception data to the management section 335. That is, in the memory data transmission process of FIG. 31, as described above, in step S223, the mobile phone 30 transmits the latest memory data together with the update information in step S223. Then, the memory data and the update information are received.
- step S 274 the management unit 3 3 5 refers to the update information received from the calling mobile phone 3 0 1 ⁇ . 0 1! It is determined whether the latest memory data for the user is stored.
- step S274 If it is determined in step S274 that the latest memory data for the user of the mobile phone 301 on the calling side has already been stored in the decoding unit 3332, the process proceeds to step S275.
- the management unit 335 discards the memory data and the update information received in step S2773, and proceeds to step S277.
- step S274 If it is determined in step S274 that the latest memory data for the user of the calling mobile phone 30 has not yet been stored in the decryption unit 3332, the flow advances to step S276.
- the management unit 335 has transmitted the latest memory data obtained in step S 273 together with the telephone number of the calling mobile phone 30 received at the time of the incoming call, and the memory data.
- the contents of the table memory 391 of the decoding unit 3332 are updated by storing the information in the decoding unit 3332 in association with the update information.
- step S277 the management unit 3335 controls the transmission control unit 3224 of the transmission unit 3113 to transmit a preparation completion notification indicating that the preparation for the voice call is completed.
- the data is transmitted as transmission data, and the flow advances to step S278.
- step S 278 the reception control section 33 1 outputs the encoded voice data included in the reception data supplied thereto to the decoding section 3 32, and is in a state where voice communication is possible, and the memory The data update processing ends.
- step S291 the frame cutout unit 371 frames the encoded voice data input from the reception control unit 331.
- step S292 the DCT conversion section 3772 performs M-DCT conversion on the encoded voice data framed in step S291, and as DCT spectrum data, the class code separation section 3772 Output to
- step S293 the class code separation unit 3733 separates the DCT spectrum data into a class code and low-frequency components, outputs the class code to the table memory 391, and outputs The band frequency component is output to the spectrum coupling section 3776.
- step S294 the table memory 3991 adjusts the gain of the DCT spectrum data of the high frequency component stored in the address corresponding to the class code output by the class code separation unit 373. Output to section 3 7 5.
- step S295 the spectrum coupling section 376 converts the high-frequency component input with the gain adjusted in the gain adjustment section 375 with the low-frequency component input from the class code separation section 373.
- the components are spectrally combined and output to the inverse DCT transformer 377.
- step S296 the inverse DCT transform section 377 performs the inverse DCT transform on the combined DCT spectrum data.
- step S297 the frame combining section 378 performs an overlap process and outputs it to the D / A conversion section 333 (FIG. 25) as a wideband digital audio signal.
- the cellular phone 3 0 1 2 Odor incoming side Te as long as the latest memory data about the user of the mobile phone 3 0 outgoing side is not stored, always stored contents Will be updated.
- the user only to request the updating of the memory data when needed, only when the request is turned on, may be performed update processing described above c
- the telephone number transmitted from the calling side at the time of an incoming call is used as the specific information for specifying the calling side.
- a unique ID (Identification) can be assigned to a user or the like, and that ID can be used as specific information.
- the operation when the memory is fixed is the memory data transmission process (the process shown in FIG. 31) in step S202 of FIG. 30 and the memory data update process (the process shown in step S252) of FIG. This is the same as the case where the processing shown in Fig. 34) is omitted.
- the above-described series of processing can be performed by hardware or can be performed by software.
- a program constituting the software is installed on a general-purpose computer or the like.
- FIG. 36 shows a configuration example of an embodiment of a computer in which a program for executing the above-described series of processes is installed.
- the program can be recorded in advance on a hard disk 405 or ROM 403 as a recording medium built in the computer.
- the program is a flexible disk, CD-ROM (Compact Disc)
- Disc a magnetic disk, a semiconductor memory, or other removable recording medium 411, which can be temporarily or permanently stored (recorded).
- a removable recording medium 411 can be provided as so-called package software.
- the program can be installed on the computer from the removable recording medium 411 as described above, or transmitted from the download site to a computer via a satellite for digital satellite broadcasting by wireless, LAN ( Local Area
- Network Via a network such as the Internet, the data is transferred to the computer by wire, and the computer transfers the program transferred in that way. It can be received by the communication unit 408 and installed on the built-in hard disk 405.
- the computer has a built-in CPU (Central Processing Unit) 402.
- CPU Central Processing Unit
- the CPU 402 is connected to an input / output interface 410 via a bus 401, and the CPU 402 is connected to the keyboard and mouse by the user via the input / output interface 410.
- a command is input by operating the input unit 407 including a microphone or the like, a program stored in a ROM (Read Only Memory) 403 is executed in accordance with the command.
- the CPU 402 is a program stored in the hard disk 405, a program transferred from a satellite or a network, received by the communication unit 408, and installed in the hard disk 405, or a drive.
- the program read from the removable recording medium 411 mounted on the 409 and installed on the hard disk 405 is loaded into a RAM (Random Access Memory) 404 and executed.
- the CPU 402 performs the processing according to the above-described flowchart or the processing performed by the configuration of the above-described block diagram. Then, the CPU 402 transmits the processing result from an output unit 406 configured by an LCD (Liquid Crystal Display), a speaker, or the like, as necessary, for example, via an input / output interface 410. It is output or transmitted from the communication unit 408, and further recorded on the hard disk 405.
- an output unit 406 configured by an LCD (Liquid Crystal Display), a speaker, or the like
- processing steps for describing a program for causing a computer to perform various kinds of processing do not necessarily need to be processed in a time series in the order described as a flowchart, and may be performed in parallel or in parallel. It also includes processes that are executed individually (for example, parallel processing or processing by objects).
- the program may be processed by one computer, or may be processed in a distributed manner by a plurality of computers.
- the program in the form of a good c present even executed by being transferred to a remote computer, the present invention, audio (image signal) broadening apparatus and a voice call by the cellular phone each other has been described in the case of applying to a transmission system that performs
- the present invention can be widely applied to systems for performing voice communication, such as improving sound quality of fixed telephone voice, AM (Amplitude modulation) / FM (Frequency Modulation) radio broadcast received voice, and analog TV (Television) broadcast received voice.
- the term “system” refers to an overall device including a plurality of devices, means, and the like. Industrial applicability
- a system capable of accurately restoring a high frequency component of a narrowband signal in which a high frequency component is suppressed is provided. Can be realized.
- the high frequency of the narrow band signal in which the high frequency component is suppressed with respect to the other signal processing devices The components can be accurately restored.
- the second signal processing device As described above, according to the second signal processing device, the signal processing method, the recording medium, and the program of the present invention, it is possible to accurately restore the high frequency component of the narrowband signal in which the high frequency component is suppressed. become.
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Computational Linguistics (AREA)
- Acoustics & Sound (AREA)
- Quality & Reliability (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Compression, Expansion, Code Conversion, And Decoders (AREA)
- Picture Signal Circuits (AREA)
- Facsimile Image Signal Circuits (AREA)
- Telephonic Communication Services (AREA)
Abstract
Description
Claims
Priority Applications (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP03795239A EP1538602B1 (en) | 2002-09-12 | 2003-08-27 | Wideband synthesis from a narrowband signal |
DE60325741T DE60325741D1 (de) | 2002-09-12 | 2003-08-27 | Breitbandsynthese von schmalbandigen signalen |
US10/495,295 US7668319B2 (en) | 2002-09-12 | 2003-08-27 | Signal processing system, signal processing apparatus and method, recording medium, and program |
US12/571,592 US7986797B2 (en) | 2002-09-12 | 2009-10-01 | Signal processing system, signal processing apparatus and method, recording medium, and program |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2002266327A JP3879922B2 (ja) | 2002-09-12 | 2002-09-12 | 信号処理システム、信号処理装置および方法、記録媒体、並びにプログラム |
JP2002/266327 | 2002-09-12 |
Related Child Applications (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US10495295 A-371-Of-International | 2003-08-27 | ||
US12/571,592 Division US7986797B2 (en) | 2002-09-12 | 2009-10-01 | Signal processing system, signal processing apparatus and method, recording medium, and program |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2004025625A1 true WO2004025625A1 (ja) | 2004-03-25 |
Family
ID=31986636
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2003/010824 WO2004025625A1 (ja) | 2002-09-12 | 2003-08-27 | 信号処理システム、信号処理装置および方法、記録媒体、並びにプログラム |
Country Status (7)
Country | Link |
---|---|
US (2) | US7668319B2 (ja) |
EP (1) | EP1538602B1 (ja) |
JP (1) | JP3879922B2 (ja) |
KR (1) | KR101001475B1 (ja) |
CN (1) | CN1302457C (ja) |
DE (1) | DE60325741D1 (ja) |
WO (1) | WO2004025625A1 (ja) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8818541B2 (en) | 2009-01-16 | 2014-08-26 | Dolby International Ab | Cross product enhanced harmonic transposition |
CN109688531A (zh) * | 2017-10-18 | 2019-04-26 | 宏达国际电子股份有限公司 | 获取高音质音频变换信息的方法、电子装置及记录介质 |
Families Citing this family (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP3879922B2 (ja) * | 2002-09-12 | 2007-02-14 | ソニー株式会社 | 信号処理システム、信号処理装置および方法、記録媒体、並びにプログラム |
EP1672618B1 (en) * | 2003-10-07 | 2010-12-15 | Panasonic Corporation | Method for deciding time boundary for encoding spectrum envelope and frequency resolution |
CN101656073B (zh) * | 2004-05-14 | 2012-05-23 | 松下电器产业株式会社 | 解码装置、解码方法以及通信终端和基站装置 |
BRPI0510014B1 (pt) * | 2004-05-14 | 2019-03-26 | Panasonic Intellectual Property Corporation Of America | Dispositivo de codificação, dispositivo de decodificação e método do mesmo |
EP1638083B1 (en) * | 2004-09-17 | 2009-04-22 | Harman Becker Automotive Systems GmbH | Bandwidth extension of bandlimited audio signals |
JP4899359B2 (ja) * | 2005-07-11 | 2012-03-21 | ソニー株式会社 | 信号符号化装置及び方法、信号復号装置及び方法、並びにプログラム及び記録媒体 |
JP5383008B2 (ja) * | 2007-07-09 | 2014-01-08 | アルパイン株式会社 | 音声明瞭度改善システム及び音声明瞭度改善方法 |
ATE477572T1 (de) * | 2007-10-01 | 2010-08-15 | Harman Becker Automotive Sys | Effiziente audiosignalverarbeitung im subbandbereich, verfahren, vorrichtung und dazugehöriges computerprogramm |
JP2009300707A (ja) * | 2008-06-13 | 2009-12-24 | Sony Corp | 情報処理装置および方法、並びにプログラム |
ES2805349T3 (es) | 2009-10-21 | 2021-02-11 | Dolby Int Ab | Sobremuestreo en un banco de filtros de reemisor combinado |
CN102436820B (zh) * | 2010-09-29 | 2013-08-28 | 华为技术有限公司 | 高频带信号编码方法及装置、高频带信号解码方法及装置 |
CN105761724B (zh) * | 2012-03-01 | 2021-02-09 | 华为技术有限公司 | 一种语音频信号处理方法和装置 |
KR101897455B1 (ko) | 2012-04-16 | 2018-10-04 | 삼성전자주식회사 | 음질 향상 장치 및 방법 |
US10008198B2 (en) * | 2013-03-28 | 2018-06-26 | Korea Advanced Institute Of Science And Technology | Nested segmentation method for speech recognition based on sound processing of brain |
KR20150032390A (ko) * | 2013-09-16 | 2015-03-26 | 삼성전자주식회사 | 음성 명료도 향상을 위한 음성 신호 처리 장치 및 방법 |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH08123484A (ja) * | 1994-10-28 | 1996-05-17 | Matsushita Electric Ind Co Ltd | 信号合成方法および信号合成装置 |
JPH08278800A (ja) * | 1995-04-05 | 1996-10-22 | Fujitsu Ltd | 音声通信システム |
JPH09101798A (ja) * | 1995-10-05 | 1997-04-15 | Matsushita Electric Ind Co Ltd | 音声帯域拡大方法および音声帯域拡大装置 |
JP2000305599A (ja) * | 1999-04-22 | 2000-11-02 | Sony Corp | 音声合成装置及び方法、電話装置並びにプログラム提供媒体 |
Family Cites Families (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2779886B2 (ja) * | 1992-10-05 | 1998-07-23 | 日本電信電話株式会社 | 広帯域音声信号復元方法 |
SE501981C2 (sv) * | 1993-11-02 | 1995-07-03 | Ericsson Telefon Ab L M | Förfarande och anordning för diskriminering mellan stationära och icke stationära signaler |
EP0732687B2 (en) * | 1995-03-13 | 2005-10-12 | Matsushita Electric Industrial Co., Ltd. | Apparatus for expanding speech bandwidth |
EP0945852A1 (en) * | 1998-03-25 | 1999-09-29 | BRITISH TELECOMMUNICATIONS public limited company | Speech synthesis |
US6658155B1 (en) * | 1999-03-25 | 2003-12-02 | Sony Corporation | Encoding apparatus |
US6351733B1 (en) * | 2000-03-02 | 2002-02-26 | Hearing Enhancement Company, Llc | Method and apparatus for accommodating primary content audio and secondary content remaining audio capability in the digital audio production process |
JP4596196B2 (ja) * | 2000-08-02 | 2010-12-08 | ソニー株式会社 | ディジタル信号処理方法、学習方法及びそれらの装置並びにプログラム格納媒体 |
JP4538705B2 (ja) * | 2000-08-02 | 2010-09-08 | ソニー株式会社 | ディジタル信号処理方法、学習方法及びそれらの装置並びにプログラム格納媒体 |
JP3879922B2 (ja) * | 2002-09-12 | 2007-02-14 | ソニー株式会社 | 信号処理システム、信号処理装置および方法、記録媒体、並びにプログラム |
-
2002
- 2002-09-12 JP JP2002266327A patent/JP3879922B2/ja not_active Expired - Fee Related
-
2003
- 2003-08-27 EP EP03795239A patent/EP1538602B1/en not_active Expired - Lifetime
- 2003-08-27 KR KR20047007114A patent/KR101001475B1/ko not_active IP Right Cessation
- 2003-08-27 CN CNB038017334A patent/CN1302457C/zh not_active Expired - Fee Related
- 2003-08-27 WO PCT/JP2003/010824 patent/WO2004025625A1/ja active Application Filing
- 2003-08-27 US US10/495,295 patent/US7668319B2/en not_active Expired - Fee Related
- 2003-08-27 DE DE60325741T patent/DE60325741D1/de not_active Expired - Lifetime
-
2009
- 2009-10-01 US US12/571,592 patent/US7986797B2/en not_active Expired - Fee Related
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH08123484A (ja) * | 1994-10-28 | 1996-05-17 | Matsushita Electric Ind Co Ltd | 信号合成方法および信号合成装置 |
JPH08278800A (ja) * | 1995-04-05 | 1996-10-22 | Fujitsu Ltd | 音声通信システム |
JPH09101798A (ja) * | 1995-10-05 | 1997-04-15 | Matsushita Electric Ind Co Ltd | 音声帯域拡大方法および音声帯域拡大装置 |
JP2000305599A (ja) * | 1999-04-22 | 2000-11-02 | Sony Corp | 音声合成装置及び方法、電話装置並びにプログラム提供媒体 |
Non-Patent Citations (3)
Title |
---|
HOSOKI M. ET AL.: "Speech signal band width extension and noise remova using subband-HMM", PROC. IEEE ICASSP'02, vol. 1, 13 May 2002 (2002-05-13) - 17 May 2002 (2002-05-17), pages I-245 - I-248, XP002975699 * |
YOSHIHISA NAKATOH ET AL.: "Spectrum senkei shazo ni yoru taiiki seigen onsei no kotaiikika", THE TRANSACTIONS OF THE INSTITUTE OF ELECTRONICS, INFORMATION AND COMMUNICATION ENGINEERS, D-II, vol. J83, no. 11, 25 November 2000 (2000-11-25), pages 2246 - 2254, XP002975697 * |
YUKI YOSHIDA ET AL.: "Code book mapping ni yoru kyotaiiki onsei kara kotaiiki onsei no seiseiho", THE TRANSACTIONS OF THE INSITUTE OF ELECTRONICS, INFORMATION AND COMMUNICATION ENGINEERS, D-II, vol. J78, no. 3, March 1995 (1995-03-01), pages 391 - 399, XP002975698 * |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8818541B2 (en) | 2009-01-16 | 2014-08-26 | Dolby International Ab | Cross product enhanced harmonic transposition |
US9799346B2 (en) | 2009-01-16 | 2017-10-24 | Dolby International Ab | Cross product enhanced harmonic transposition |
US10192565B2 (en) | 2009-01-16 | 2019-01-29 | Dolby International Ab | Cross product enhanced harmonic transposition |
US10586550B2 (en) | 2009-01-16 | 2020-03-10 | Dolby International Ab | Cross product enhanced harmonic transposition |
US11031025B2 (en) | 2009-01-16 | 2021-06-08 | Dolby International Ab | Cross product enhanced harmonic transposition |
US11682410B2 (en) | 2009-01-16 | 2023-06-20 | Dolby International Ab | Cross product enhanced harmonic transposition |
US11935551B2 (en) | 2009-01-16 | 2024-03-19 | Dolby International Ab | Cross product enhanced harmonic transposition |
US12119011B2 (en) | 2009-01-16 | 2024-10-15 | Dolby International Ab | Cross product enhanced harmonic transposition |
CN109688531A (zh) * | 2017-10-18 | 2019-04-26 | 宏达国际电子股份有限公司 | 获取高音质音频变换信息的方法、电子装置及记录介质 |
Also Published As
Publication number | Publication date |
---|---|
KR101001475B1 (ko) | 2010-12-14 |
JP2004102095A (ja) | 2004-04-02 |
EP1538602A4 (en) | 2007-07-18 |
US20050073986A1 (en) | 2005-04-07 |
EP1538602A1 (en) | 2005-06-08 |
CN1602516A (zh) | 2005-03-30 |
KR20050037431A (ko) | 2005-04-21 |
EP1538602B1 (en) | 2009-01-07 |
DE60325741D1 (de) | 2009-02-26 |
US7668319B2 (en) | 2010-02-23 |
CN1302457C (zh) | 2007-02-28 |
JP3879922B2 (ja) | 2007-02-14 |
US20100020827A1 (en) | 2010-01-28 |
US7986797B2 (en) | 2011-07-26 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US7986797B2 (en) | Signal processing system, signal processing apparatus and method, recording medium, and program | |
US10559313B2 (en) | Speech/audio signal processing method and apparatus | |
US8688440B2 (en) | Coding apparatus, decoding apparatus, coding method and decoding method | |
US6539355B1 (en) | Signal band expanding method and apparatus and signal synthesis method and apparatus | |
JP5933965B2 (ja) | 高周波数の再構成方法を使用するコーディング・システムの性能拡大方法 | |
JP3579047B2 (ja) | オーディオ復号装置と復号方法およびプログラム | |
RU2383943C2 (ru) | Кодирование звуковых сигналов | |
WO2013027629A1 (ja) | 符号化装置および方法、復号装置および方法、並びにプログラム | |
US20100138219A1 (en) | Coding Apparatus and Decoding Apparatus | |
WO2006049204A1 (ja) | 符号化装置、復号化装置、符号化方法及び復号化方法 | |
JP2000305599A (ja) | 音声合成装置及び方法、電話装置並びにプログラム提供媒体 | |
KR20070070189A (ko) | 음성 부호화 장치 및 음성 부호화 방법 | |
JP2005107255A (ja) | サンプリングレート変換装置、符号化装置、および復号化装置 | |
JP6073456B2 (ja) | 音声強調装置 | |
JP2015507764A (ja) | オーディオ・データを処理するための方法、装置、及びシステム | |
JP2007187905A (ja) | 信号符号化装置及び方法、信号復号装置及び方法、並びにプログラム及び記録媒体 | |
KR20060131793A (ko) | 음성ㆍ악음 부호화 장치 및 음성ㆍ악음 부호화 방법 | |
US20020173969A1 (en) | Method for decompressing a compressed audio signal | |
JP2006018023A (ja) | オーディオ信号符号化装置、および符号化プログラム | |
EP1136986A2 (en) | Audio datastream transcoding apparatus | |
JP5031006B2 (ja) | スケーラブル復号化装置及びスケーラブル復号化方法 | |
JP2003186499A (ja) | 符号化装置及び復号化装置 | |
JP3594829B2 (ja) | Mpegオーディオの復号化方法 | |
JP2002208860A (ja) | データ圧縮装置とそのデータ圧縮方法及びデータ圧縮用プログラムを記録したコンピュータ読み取り可能な記録媒体、並びにデータ伸長装置とそのデータ伸長方法 | |
JP3504485B2 (ja) | 楽音符号化装置および楽音復号化装置および楽音符号化復号化装置およびプログラム記憶媒体 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AK | Designated states |
Kind code of ref document: A1 Designated state(s): CN KR US |
|
AL | Designated countries for regional patents |
Kind code of ref document: A1 Designated state(s): AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IT LU MC NL PT RO SE SI SK TR |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2003795239 Country of ref document: EP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 1020047007114 Country of ref document: KR |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application | ||
WWE | Wipo information: entry into national phase |
Ref document number: 20038017334 Country of ref document: CN |
|
WWE | Wipo information: entry into national phase |
Ref document number: 10495295 Country of ref document: US |
|
WWP | Wipo information: published in national office |
Ref document number: 2003795239 Country of ref document: EP |