Speech/audio bitstream decoding method and apparatus

• US 10,269,357 B2
• Filed: 09/02/2016
• Issued: 04/23/2019
• Est. Priority Date: 03/21/2014
• Status: Active Grant

• Current Assignee: Huawei Technologies Co., Ltd. (Huawei Investment & Holding Co., Ltd.)
• Original Assignee: Huawei Technologies Co., Ltd. (Huawei Investment & Holding Co., Ltd.)
• Inventors: Zhang, Xingtao, Liu, Zexin, Miao, Lei
• Primary Examiner(s): Chawan, Vijay B
• Application Number:
  US15/256,018
• Publication Number:
  US 20160372122A1
• Time in Patent Office: 963 Days
• Field of Search: 704219, 704500-504, 704230, 704229, 7042001, 704203, 704207, 704226, 704201, 704208, 704211, 704220, 704221, 704223, 704225, 704228
• CPC Class Codes: G10L 19/005 Correction of errors induced by the transmission channel, if related to the coding algorithm
• CPC Class Codes: G10L 19/06 Determination or coding of the spectral characteristics, e.g. of the short-term prediction coefficients
• CPC Class Codes: G10L 19/167 Audio streaming, i.e. formatting and decoding of an encoded audio signal representation into a data stream for transmission or storage purposes
• CPC Class Codes: G10L 19/26 Pre-filtering or post-filtering
• CPC Class Codes: G10L 2019/0002 Codebook adaptations

Abstract

The present invention disclose a speech/audio bitstream decoding method including: acquiring a speech/audio decoding parameter of a current speech/audio frame, where the foregoing current speech/audio frame is a redundant decoded frame or a speech/audio frame previous to the foregoing current speech/audio frame is a redundant decoded frame; performing post processing on the acquired speech/audio decoding parameter according to speech/audio parameters of X speech/audio frames, where the foregoing X speech/audio frames include M speech/audio frames previous to the foregoing current speech/audio frame and/or N speech/audio frames next to the foregoing current speech/audio frame; and recovering a speech/audio signal by using the post-processed speech/audio decoding parameter of the foregoing current speech/audio frame. The technical solutions of the present invention help improve quality of an output speech/audio signal.

Claims

1. An audio bitstream decoding method implemented by a decoder, comprising:

acquiring, by a network interface of the decoder, a decoding parameter of a frame from an input audio bitstream, wherein the frame is a redundant decoded frame that is recovered based on redundant bitstream information from another frame when the frame is a lost frame, or a previous frame adjacent to the frame that is a redundant decoded frame, and the decoding parameter comprises an adaptive codebook gain;

adjusting, by a processor of the decoder, the adaptive codebook gain of the frame according to a signal class, an algebraic codebook gain, or an adaptive codebook gain of X frames of the audio bitstream, to obtain an adjusted adaptive codebook gain of the frame, wherein the X frames comprise M frames previous to the frame and/or N frames next to the frame, and wherein X, M and N are positive integers;

recovering, by the processor of the decoder, a signal of the frame according to the adjusted adaptive codebook gain of the frame; and

outputting an audio signal synthesized according to the recovered signal.

2. The method according to claim 1, wherein adjusting the adaptive codebook gain comprises:

attenuating an adaptive codebook gain of a subframe of the frame, wherein the frame is a redundant decoded frame, a signal class of the frame is not unvoiced, a signal class of at least one of two frames next to the frame is unvoiced, and an algebraic codebook gain of the subframe is greater than or equal to an algebraic codebook gain of the previous frame adjacent to the frame.

3. The method according to claim 1, wherein adjusting the adaptive codebook gain comprises:

attenuating an adaptive codebook gain of a subframe of the frame, wherein the frame is a redundant decoded frame, the signal class of the frame is not unvoiced, the signal class of at least one of two frames next to the frame is unvoiced, and the algebraic codebook gain of the subframe is greater than or equal to an algebraic codebook gain of a previous subframe adjacent to the subframe.

4. The method according to claim 1, wherein the decoding parameter of the frame further comprises an algebraic codebook, and wherein the method further comprises:

performing post processing on the algebraic codebook of the frame according to a signal class, an algebraic codebook, or a spectrum tilt factor of the X frames, to obtain a post-processed algebraic codebook of the frame.

5. The method according to claim 1, wherein the decoding parameter of the frame further comprises a bandwidth extension envelope, and wherein the method further comprises:

performing post processing on the bandwidth extension envelope of the frame according to a signal class, a bandwidth extension envelope, or a spectrum tilt factor of the X frames, to obtain a post-processed bandwidth extension envelope of the frame.

6. The method according to claim 5,

wherein the previous frame adjacent to the frame is a normal decoded frame, a signal class of the previous frame adjacent to the frame is the same as that of a next frame adjacent to the frame, and wherein the performing post processing on the bandwidth extension envelope of the frame comprises:

obtaining the post-processed bandwidth extension envelope of the frame based on a bandwidth extension envelope of the previous frame adjacent to the frame and the bandwidth extension envelope of the frame.

7. The method according to claim 6, wherein

a signal class of the frame is not unvoiced, a signal class of the next frame adjacent to the frame is unvoiced, and a spectrum tilt factor of the previous frame adjacent to the frame is less than or equal to a tenth threshold, and the method further comprises:

modifying the bandwidth extension envelope of the frame according to the bandwidth extension envelope or the spectrum tilt factor of the previous frame adjacent to the frame, to obtain the post-processed bandwidth extension envelope of the frame.

8. The method according to claim 7, wherein a modification factor for modifying the bandwidth extension envelope of the frame is inversely proportional to the spectrum tilt factor of the previous frame adjacent to the frame, and is proportional to a ratio of the bandwidth extension envelope of the previous frame adjacent to the frame to the bandwidth extension envelope of the frame.

9. The method according claim 1, wherein the decoding parameter of the frame further comprises a pitch period, and wherein the method further comprises:

performing post processing on the pitch period of the frame according to the signal class or a pitch period of the X frames, to obtain a post-processed pitch period of the frame.

10. A decoder for decoding an audio bitstream, comprising:

a memory storing instructions, and a processor coupled to the memory to executes the instructions, the processor configured to;

acquire, via an interface, a decoding parameter of a frame from the audio bitstream, wherein the frame is a redundant decoded frame that is recovered based on redundant bitstream information from another frame when the frame is a lost frame, or a previous frame adjacent to the frame that is a redundant decoded frame, and wherein the decoding parameter comprises an adaptive codebook gain;

adjust the adaptive codebook gain of the frame according to a signal class, an algebraic codebook gain, or an adaptive codebook gain of X frames of the audio bitstream when the frame is a redundant decoded frame or a previous frame adjacent to the frame is a redundant decoded frame, to obtain an adjusted adaptive codebook gain of the frame, wherein the X frames comprise M frames previous to the frame and/or N frames next to the frame, and wherein X, M and N are positive integers;

recover a signal of the frame according to the adjusted adaptive codebook gain of the frame; and

output, via the interface, an audio signal synthesized according to the recovered signal.

11. The decoder according to claim 10, wherein the processor is configured to:

attenuating an adaptive codebook gain of a subframe of the frame when the frame is a redundant decoded frame, a signal class of the frame is not unvoiced, a signal class of at least one of two frames next to the frame is unvoiced, and an algebraic codebook gain of the subframe is greater than or equal to an algebraic codebook gain of the previous frame adjacent to the frame.

12. The decoder according to claim 10, wherein the processor is further configured to:

attenuate an adaptive codebook gain of a subframe of the frame when the frame is a redundant decoded frame, the signal class of the frame is not unvoiced, the signal class of at least one of two frames next to the frame is unvoiced, and the algebraic codebook gain of the subframe is greater than or equal to an algebraic codebook gain of a previous subframe adjacent to the subframe.

13. The decoder according to claim 10, wherein the decoding parameter of the frame further comprises a bandwidth extension envelope, and the processor is further configured to:

perform post processing on the bandwidth extension envelope of the frame to obtain a post-processed bandwidth extension envelope of the frame, wherein the post processing is performed according to a signal class, a bandwidth extension envelope, or a spectrum tilt factor of the X frames.

14. The decoder according to claim 10, wherein the processor is configured to:

obtain the post-processed bandwidth extension envelope of the frame when the previous frame adjacent to the frame is a normal decoded frame, and the signal class of the previous frame adjacent to the frame is the same as that of a next frame adjacent to the frame, wherein the post-processed bandwidth extension envelope of the frame is obtained based on a bandwidth extension envelope of the previous frame adjacent to the frame and the bandwidth extension envelope of the frame.

15. The decoder according to claim 14, wherein the processor is further configured to:

modify the bandwidth extension envelope of the frame when a signal class of the frame is not unvoiced, a signal class of the next frame adjacent to the frame is unvoiced, and a spectrum tilt factor of the previous frame adjacent to the frame is less than or equal to a tenth threshold, wherein the bandwidth extension envelope of the frame is modified according to the bandwidth extension envelope or the spectrum tilt factor of the previous frame adjacent to the frame, to obtain the post-processed bandwidth extension envelope of the frame.

16. The decoder according to claim 15, wherein a modification factor used by the processor for modifying the bandwidth extension envelope of the frame is inversely proportional to the spectrum tilt factor of the previous frame adjacent to the frame, and is proportional to a ratio of the bandwidth extension envelope of the previous frame adjacent to the frame to the bandwidth extension envelope of the frame.

17. The decoder according to claim 10, wherein the decoding parameter of the frame further comprises a pitch period, and the processor is further configured to:

perform post processing on the pitch period of the frame according to at least one of the signal class or a pitch period of the X frames, to obtain a post-processed pitch period of the frame.