SIGNAL CLASSIFICATION METHOD AND DEVICE, AND ENCODING AND DECODING METHODS AND DEVICES

• US 20130117029A1
• Filed: 12/27/2012
• Published: 05/09/2013
• Est. Priority Date: 05/25/2011
• 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: LIU, Zexin, MIAO, Lei, TALEB, Anisse
• Granted Patent: US 8,600,765 B2
• Time in Patent Office: Days
• US Class Current: 704/500
• CPC Class Codes: G10L 19/008 Multichannel audio signal coding or decoding using interchannel correlation to reduce redundancy, e.g. joint-stereo, intensity-coding or matrixing
• CPC Class Codes: G10L 19/0204 using subband decomposition
• CPC Class Codes: G10L 19/20 using sound class specific coding, hybrid encoders or object based coding
• CPC Class Codes: G10L 19/265 Pre-filtering, e.g. high frequency emphasis prior to encoding
• CPC Class Codes: G10L 21/0388 Details of processing therefor

Abstract

Embodiments of the present invention provide a signal classification method and device, and encoding and decoding methods and devices. The encoding method includes: dividing a current frame into a low-frequency band signal and a high-frequency band signal; attenuating the high-frequency band signal or a to-be-encoded characteristic parameter of the high-frequency band signal according to an energy attenuation value of the low-frequency band signal, where the energy attenuation value indicates energy attenuation of the low-frequency band signal caused by encoding of the low-frequency band signal; and encoding the attenuated high-frequency band signal or the attenuated to-be-encoded characteristic parameter of the high-frequency band signal. The technical solutions according to the embodiments of the present invention can improve the effect of combining the low-frequency band signal and the high-frequency band signal at the decoder.

Claims

1. An encoding method, comprising:

dividing a current frame into a low-frequency band signal and a high-frequency band signal;

attenuating a one of the group consisting of the high-frequency band signal and a to-be-encoded characteristic parameter of the high-frequency band signal, the attenuating being according to an energy attenuation value of the low-frequency band signal, and wherein the energy attenuation value indicates energy attenuation of the low-frequency band signal caused by encoding of the low-frequency band signal; and

encoding the one of the group consisting of the attenuated high-frequency band signal and the attenuated to-be-encoded characteristic parameter of the high-frequency band signal.

2. The method according to claim 1, further comprising:

determining a signal class of the high-frequency band signal; and

wherein the attenuating the one of the group consisting of the high-frequency band signal and the to-be-encoded characteristic parameter of the high-frequency band signal according to the energy attenuation value of the low-frequency band signal comprises;

attenuating the one of the group consisting of the high-frequency band signal and the to-be-encoded characteristic parameter of the high-frequency band signal according to the energy attenuation value and the signal class of the high-frequency band signal.

3. The method according to claim 2, wherein

the attenuating the one of the group consisting of the high-frequency band signal and the to-be-encoded characteristic parameter of the high-frequency band signal according to the energy attenuation value and the signal class of the high-frequency band signal comprises conditionally executing the following:

when the signal class of the high-frequency band signal is a transient class, attenuating a one of the group consisting of a high-frequency band time-domain signal and a to-be-encoded time domain envelope of the high-frequency band signal according to the energy attenuation value;

and,

when the signal class of the high-frequency band signal is a one of the group consisting of a fricative class, a harmonic class and a normal class, attenuating one of the group consisting of a high-frequency band frequency-domain signal and a to-be-encoded frequency domain envelope of the high-frequency band signal according to the energy attenuation value.

4. The method according to claim 1, further comprising:

encoding the low-frequency band signal and locally decoding a result of encoding the low-frequency band signal; and

using a ratio of energy of the low-frequency band signal to energy of a signal obtained by the locally decoding as the energy attenuation value.

5. The method according to claim 1, wherein

the energy attenuation value is a preset value, and the energy attenuation value is obtained according to ratios of energy of multiple low-frequency band signals of a same-class frame to energy of signals obtained by decoding encoding results of the low-frequency band signals of the same-class frame, wherein the same-class frame is a data frame of the same signal class as the high-frequency band signal of the current frame.

6. A decoding method, comprising:

decoding a bit stream to obtain a one of the group consisting of (a) a high-frequency band signal of a current frame and (b) a characteristic parameter of the high-frequency band signal of the current frame; and

attenuating the one of the group consisting of (a) the high-frequency band signal and (b) the characteristic parameter of the high-frequency band signal according to an energy attenuation value of a low-frequency band signal of the current frame, wherein the energy attenuation value indicates energy attenuation of the low-frequency band signal caused by encoding of the low-frequency band signal.

7. The method according to claim 6, further comprising:

decoding the bit stream to obtain a signal class of the high-frequency band signal of the current frame; and

wherein the attenuating the one of the group consisting of the high-frequency band signal and the characteristic parameter of the high-frequency band signal according to the energy attenuation value of the low-frequency band signal of the current frame comprises;

attenuating the one of the group consisting of the high-frequency band signal and the characteristic parameter of the high-frequency band signal according to the energy attenuation value and the signal class of the high-frequency band signal of the current frame.

8. The method according to claim 7, wherein

the attenuating the one of the group consisting of the high-frequency band signal and the characteristic parameter of the high-frequency band signal according to the energy attenuation value and the signal class of the high-frequency band signal of the current frame comprises conditionally executing the following:

when the signal class of the high-frequency band signal of the current frame is one of the group consisting of a transient class, attenuating a high-frequency band time-domain signal and a time-domain envelope of the high-frequency band signal according to the energy attenuation value;

and

when the signal class of the high-frequency band signal of the current frame is one of the group consisting of a fricative class, a harmonic class and a normal class, attenuating one of the group consisting of a high-frequency band frequency-domain signal and a frequency-domain envelope of the high-frequency band signal according to the energy attenuation value.

9. The method according to claim 6, further comprising:

decoding the bit stream to obtain the energy attenuation value, wherein the energy attenuation value indicates;

a ratio of energy of the low-frequency band signal of the current frame to energy of a signal obtained by locally decoding a result of encoding, by an encoder, the low-frequency band signal of the current frame.

10. The method according to claim 6, wherein

the energy attenuation value is a preset value, and the energy attenuation value is obtained according to a ratio of energy of a low-frequency band signal of a same-class frame to energy of a signal obtained by decoding a result of encoding the low-frequency band signal of the same-class frame, wherein the same-class frame is a data frame of the same signal class as the high-frequency band signal of the current frame.

11. An encoding device, comprising:

a division unit, configured to divide a current frame into a low-frequency band signal and a high-frequency band signal;

a correction unit, configured to attenuate a one of the group consisting of the high-frequency band signal and a to-be-encoded characteristic parameter of the high-frequency band signal according to an energy attenuation value of the low-frequency band signal, wherein the energy attenuation value indicates energy attenuation of the low-frequency band signal caused by encoding of the low-frequency band signal of the current frame; and

an encoding unit, configured to encode the one of the group consisting of the attenuated high-frequency band signal and the attenuated to-be-encoded characteristic parameter of the high-frequency band signal.

12. The device according to claim 11, further comprising:

a signal class determining unit, configured to determine a signal class of the high-frequency band signal;

wherein the correction unit is configured to attenuate the one of the group consisting of the high-frequency band signal and the to-be-encoded characteristic parameter of the high-frequency band signal according to the energy attenuation value and the signal class of the high-frequency band signal.

13. The device according to claim 12, wherein

the correction unit is configured to conditionally:

attenuate the one of the group consisting of a high-frequency band time-domain signal and a to-be-encoded time domain envelope of the high-frequency band signal according to the energy attenuation value when the signal class of the high-frequency band signal is a transient class;

and

attenuate the one of the group consisting of a high-frequency band frequency-domain signal and a to-be-encoded frequency domain envelope of the high-frequency band signal according to the energy attenuation value when the signal class of the high-frequency band signal is one of the group consisting of a fricative class, a harmonic class and a normal class.

14. The device according to claim 11, further comprises:

an energy attenuation value obtaining unit configured to encode the low-frequency band signal and locally decode a result of encoding the low-frequency band signal; and

use a ratio of energy of the low-frequency band signal to energy of a signal obtained by the local decoding as the energy attenuation value.

15. The device according to claim 11, further comprises:

an energy attenuation value setting unit configured to set the energy attenuation value, wherein the energy attenuation value is obtained according to ratios of energy of multiple low-frequency band signals of a same-class frame to energy of signals obtained by decoding results of encoding the low-frequency band signals of the same-class frame, wherein the same-class frame is a data frame of the same signal class as the high-frequency band signal of the current frame.

16. A decoding device, comprising:

a decoding unit configured to decode a bit stream to obtain a one of the group consisting of a high-frequency band signal of a current frame and a characteristic parameter of the high-frequency band signal of the current frame; and

a correction unit, configured to attenuate the one of the group consisting of the high-frequency band signal and the characteristic parameter of the high-frequency band signal according to an energy attenuation value of a low-frequency band signal of the current frame, wherein the energy attenuation value indicates energy attenuation of the low-frequency band signal caused by encoding of the low-frequency band signal of the current frame.

17. The device according to claim 16, wherein

the decoding unit is further configured to decode the bit stream to obtain a signal class of the high-frequency band signal of the current frame;

and

the correction unit is configured to attenuate the one of the group consisting of the high-frequency band signal and the characteristic parameter of the high-frequency band signal according to the energy attenuation value and the signal class of the high-frequency band signal of the current frame.

18. The device according to claim 17, wherein

the correction unit is configured to conditionally execute the following:

attenuating one of the group consisting of a high-frequency band time-domain signal and a time-domain envelope of the high-frequency band signal according to the energy attenuation value when the signal class of the high-frequency band signal of the current frame is a transient class;

and

attenuating one of the group consisting of a high-frequency band frequency-domain signal and a frequency-domain envelope of the high-frequency band signal according to the energy attenuation value when the signal class of the high-frequency band signal of the current frame is one of the group consisting of a fricative class, a harmonic class and a normal class.

19. The device according to claim 16, wherein

the decoding unit is further configured to decode the bit stream to obtain the energy attenuation value, wherein the energy attenuation value indicates:

a ratio of energy of the low-frequency band signal of the current frame to energy of a signal obtained by locally decoding a result of encoding, by an encoder, the low-frequency band signal of the current frame.

20. The device according to claim 16, further comprising:

an energy attenuation value setting unit configured to set the energy attenuation value of the current frame, wherein the energy attenuation value is obtained according to a ratio of energy of a low-frequency band signal of a same-class frame to energy of a signal obtained by decoding a result of encoding the low-frequency band signal of the same-class frame, and the same-class frame is a data frame of the same signal class as the high-frequency band signal of the current frame.

21. A signal classification method, comprising:

dividing a current frame into a low-frequency band signal and a high-frequency band signal;

determining, according to a value requirement of a preset encoding/decoding characteristic parameter corresponding to a signal class, whether an encoding/decoding characteristic parameter of the current frame corresponding to the signal class meets the value requirement of the encoding/decoding characteristic parameter; and

determining a signal class of the high-frequency band signal of the current frame according to a determining result.

22. The method according to claim 21, wherein

the preset encoding/decoding characteristic parameter corresponding to the signal class comprises an encoding/decoding characteristic parameter corresponding to a noise class;

and

the encoding/decoding characteristic parameter corresponding to the noise class is one of the group consisting of;

a correlation parameter between an amplitude of a low-frequency band frequency-domain signal and an amplitude of a high-frequency band frequency-domain signal, and a correlation parameter between energy of the low-frequency band frequency-domain signal and energy of the high-frequency band frequency-domain signal.

23. The method according to claim 22, further comprising:

determining whether the number of sub-bands having a peak-to-average ratio smaller than a second threshold is greater than a second predetermined number; and

wherein the determining the signal class of the high-frequency band signal of the current frame comprises;

when the number of sub-bands having a peak-to-average ratio smaller than the second threshold is greater than the second predetermined number, and a value of the encoding/decoding characteristic parameter of the current frame corresponding to the noise class meets the value requirement of the preset encoding/decoding characteristic parameter corresponding to the noise class, determining that the signal class of the high-frequency band signal of the current frame is the noise class.

24. The method according to claim 21, wherein

the preset encoding/decoding characteristic parameter corresponding to the signal class comprises a one of the group consisting of an encoding/decoding characteristic parameter corresponding to a predicted class, and an encoding/decoding characteristic parameter corresponding to a harmonic class;

and

the encoding/decoding characteristic parameter corresponding to the predicted class and the encoding/decoding characteristic parameter corresponding to the harmonic class are one of the group consisting of;

a correlation parameter between a frequency-domain coefficient of the low-frequency band signal and a frequency-domain coefficient of the high-frequency band signal, a correlation parameter between an absolute value of the frequency-domain coefficient of the low-frequency band signal and an absolute value of the frequency-domain coefficient of the high-frequency band signal, a correlation parameter between a frequency-domain coefficient of a low-frequency band excitation spectrum and a frequency-domain coefficient of a high-frequency band excitation spectrum, and a correlation parameter between an absolute value of the frequency-domain coefficient of the low-frequency band excitation spectrum and an absolute value of the frequency-domain coefficient of the high-frequency band excitation spectrum.

25. The method according to claim 24, further comprising:

determining whether the number of sub-bands having a peak-to-average ratio greater than a first threshold is greater than a first predetermined number; and

when the preset encoding/decoding characteristic parameter corresponding to the signal class comprises the encoding/decoding characteristic parameter corresponding to the harmonic class, the determining the signal class of the high-frequency band signal of the current frame comprises;

when the number of sub-bands having a peak-to-average ratio greater than the first threshold is greater than the first predetermined number and a value of the encoding/decoding characteristic parameter of the current frame corresponding to the harmonic class meets the value requirement of the preset encoding/decoding characteristic parameter corresponding to the harmonic class, determining that the signal class of the high-frequency band signal of the current frame is the harmonic class.

26. The method according to claim 24, further comprising:

determining whether the number of sub-bands having a peak-to-average ratio greater than a first threshold is greater than a first predetermined number; and

when the preset encoding/decoding characteristic parameter corresponding to the signal class comprises the encoding/decoding characteristic parameter corresponding to the predicted class, the determining the signal class of the high-frequency band signal of the current frame comprises;

when the number of sub-bands having a peak-to-average ratio greater than the first threshold is not greater than the first predetermined number and a value of the encoding/decoding characteristic parameter of the current frame corresponding to the predicted class meets the value requirement of the preset encoding/decoding characteristic parameter corresponding to the predicted class, determining that the signal class of the high-frequency band signal of the current frame is the predicted class.

27. The method according to claim 21, wherein

the preset encoding/decoding characteristic parameter corresponding to the signal class further comprises an encoding/decoding characteristic parameter corresponding to a predicted class, and an encoding/decoding characteristic parameter corresponding to a harmonic class;

each of the encoding/decoding characteristic parameter corresponding to the predicted class and the encoding/decoding characteristic parameter corresponding to the harmonic class is one of the group consisting of;

a correlation parameter between a frequency-domain coefficient of the low-frequency band signal and a frequency-domain coefficient of the high-frequency band signal, a correlation parameter between an absolute value of the frequency-domain coefficient of the low-frequency band signal and an absolute value of the frequency-domain coefficient of the high-frequency band signal, a correlation parameter between a frequency-domain coefficient of a low-frequency band excitation spectrum and a frequency-domain coefficient of a high-frequency band excitation spectrum, and a correlation parameter between an absolute value of the frequency-domain coefficient of the low-frequency band excitation spectrum and an absolute value of the frequency-domain coefficient of the high-frequency band excitation spectrum; and

the method further comprises;

determining whether the number of sub-bands having a peak-to-average ratio greater than a first threshold is greater than a first predetermined number; and

determining whether the number of sub-bands having a peak-to-average ratio smaller than a second threshold is greater than a second predetermined number; and

wherein the determining the signal class of the high-frequency band signal of the current frame comprises conditionally executing the following;

when the number of sub-bands having a peak-to-average ratio greater than the first threshold is greater than the first predetermined number, and a value of the encoding/decoding characteristic parameter of the current frame corresponding to the harmonic class meets the value requirement of the preset encoding/decoding characteristic parameter corresponding to the harmonic class, determining that the signal class of the high-frequency band signal of the current frame is the harmonic class; and

when the number of sub-bands having a peak-to-average ratio greater than the first threshold is not greater than the first predetermined number, the number of sub-bands having a peak-to-average ratio smaller than the second threshold is not greater than the second predetermined number, and a value of the encoding/decoding characteristic parameter of the current frame corresponding to the predicted class meets the value requirement of the preset encoding/decoding characteristic parameter corresponding to the predicted class, determining that the signal class of the high-frequency band signal of the current frame is the predicted class.

28. The method according to claim 21, further comprising:

dividing a full-frequency time-domain signal of the current frame into N sub-frames, and when energy of one sub-frame is greater than a predetermined number of times of energy of a previous sub-frame of the sub-frame, determining that the signal class of the high-frequency band signal of the current frame is a transient class.

29. A signal classification device, comprising:

a division unit, configured to divide a current frame into a low-frequency band signal and a high-frequency band signal;

a judgment unit, configured to determine, according to a value requirement of a preset encoding/decoding characteristic parameter corresponding to a signal class, whether an encoding/decoding characteristic parameter of the current frame corresponding to the signal class meets the value requirement of the encoding/decoding characteristic parameter; and

a determination unit, configured to determine a signal class of the high-frequency band signal of the current frame according to a determining result.

30. The device according to claim 29, wherein

the preset encoding/decoding characteristic parameter corresponding to the signal class comprises:

an encoding/decoding characteristic parameter corresponding to a noise class; and

the encoding/decoding characteristic parameter corresponding to the noise class is one of the group consisting of;

a correlation parameter between an amplitude of a low-frequency band frequency-domain signal and an amplitude of a high-frequency band frequency-domain signal, and a correlation parameter between energy of the low-frequency band frequency-domain signal and energy of the high-frequency band frequency-domain signal.

31. The device according to claim 30, wherein

the device further comprises:

a second peak-to-average ratio judgment unit, configured to determine whether the number of sub-bands having a peak-to-average ratio smaller than a second threshold in the high-frequency band signal of the current frame is greater than a second predetermined number; and

the determination unit comprises;

a noise class determining unit, configured to determine that the signal class of the high-frequency band signal of the current frame is the noise class, when the number of sub-bands having a peak-to-average ratio smaller than the second threshold is greater than the second predetermined number, and a value of the encoding/decoding characteristic parameter of the current frame corresponding to the noise class meets the value requirement of the preset encoding/decoding characteristic parameter corresponding to the noise class.

32. The device according to claim 29, wherein

the preset encoding/decoding characteristic parameter corresponding to the signal class comprises an encoding/decoding characteristic parameter corresponding to a predicted class, or an encoding/decoding characteristic parameter corresponding to a harmonic class;

and

each of the encoding/decoding characteristic parameter corresponding to the predicted class and the encoding/decoding characteristic parameter corresponding to the harmonic class is one of the group consisting of;

a correlation parameter between a frequency-domain coefficient of the low-frequency band signal and a frequency-domain coefficient of the high-frequency band signal, a correlation parameter between an absolute value of the frequency-domain coefficient of the low-frequency band signal and an absolute value of the frequency-domain coefficient of the high-frequency band signal, a correlation parameter between a frequency-domain coefficient of a low-frequency band excitation spectrum and a frequency-domain coefficient of a high-frequency band excitation spectrum, and a correlation parameter between an absolute value of the frequency-domain coefficient of the low-frequency band excitation spectrum and an absolute value of the frequency-domain coefficient of the high-frequency band excitation spectrum.

33. The device according to claim 29, wherein

the device further comprises:

a first peak-to-average ratio judgment unit, configured to determine whether the number of sub-bands having a peak-to-average ratio greater than a first threshold in the high-frequency band signal of the current frame is greater than a first predetermined number; and

when the preset encoding/decoding characteristic parameter corresponding to the signal class comprises the encoding/decoding characteristic parameter corresponding to the harmonic class, the determination unit comprises;

a harmonic class determining unit, configured to determine that the signal class of the high-frequency band signal of the current frame is the harmonic class, when the number of sub-bands having a peak-to-average ratio greater than the first threshold is greater than the first predetermined number, and a value of the encoding/decoding characteristic parameter of the current frame corresponding to the harmonic class meets the value requirement of the preset encoding/decoding characteristic parameter corresponding to the harmonic class.

34. The device according to claim 29, wherein

the device further comprises:

a first peak-to-average ratio judgment unit, configured to determine whether the number of sub-bands having a peak-to-average ratio greater than a first threshold in the high-frequency band signal of the current frame is greater than a first predetermined number; and

when the preset encoding/decoding characteristic parameter corresponding to the signal class comprises the encoding/decoding characteristic parameter corresponding to the predicted class, the determination unit comprises;

a predicted class determining unit, configured to determine that the signal class of the high-frequency band signal of the current frame is the predicted class, when the number of sub-bands having a peak-to-average ratio greater than the first threshold is not greater than the first predetermined number, and a value of the encoding/decoding characteristic parameter of the current frame corresponding to the predicted class meets the value requirement of the preset encoding/decoding characteristic parameter corresponding to the predicted class.

35. The device according to claim 29, wherein

the preset encoding/decoding characteristic parameter corresponding to the signal class comprises an encoding/decoding characteristic parameter corresponding to a predicted class, and an encoding/decoding characteristic parameter corresponding to a harmonic class;

each of the encoding/decoding characteristic parameter corresponding to the predicted class and the encoding/decoding characteristic parameter corresponding to the harmonic class is one of the group consisting of;

a correlation parameter between a frequency-domain coefficient of the low-frequency band signal and a frequency-domain coefficient of the high-frequency band signal, a correlation parameter between an absolute value of the frequency-domain coefficient of the low-frequency band signal and an absolute value of the frequency-domain coefficient of the high-frequency band signal, a correlation parameter between a frequency-domain coefficient of a low-frequency band excitation spectrum and a frequency-domain coefficient of a high-frequency band excitation spectrum, and a correlation parameter between an absolute value of the frequency-domain coefficient of the low-frequency band excitation spectrum and an absolute value of the frequency-domain coefficient of the high-frequency band excitation spectrum;

the device further comprises;

a first peak-to-average ratio judgment unit, configured to determine whether the number of sub-bands having a peak-to-average ratio greater than a first threshold in the high-frequency band signal of the current frame is greater than a first predetermined number; and

a second peak-to-average ratio judgment unit, configured to determine whether the number of sub-bands having a peak-to-average ratio smaller than a second threshold in the high-frequency band signal of the current frame is greater than a second predetermined number; and

the determination unit comprises;

a harmonic class determining unit, configured to determine that the signal class of the high-frequency band signal of the current frame is the harmonic class, when the number of sub-bands having a peak-to-average ratio greater than the first threshold is greater than the first predetermined number, and a value of the encoding/decoding characteristic parameter of the current frame corresponding to the harmonic class meets the value requirement of the preset encoding/decoding characteristic parameter corresponding to the harmonic class; and

a predicted class determining unit, configured to determine that the signal class of the high-frequency band signal of the current frame is the predicted class, when the number of sub-bands having a peak-to-average ratio greater than the first threshold is greater than the first predetermined number, the number of sub-bands having a peak-to-average ratio smaller than the second threshold is not greater than the second predetermined number, and a value of the encoding/decoding characteristic parameter of the current frame corresponding to the predicted class meets the value requirement of the preset encoding/decoding characteristic parameter corresponding to the predicted class.

36. The device according to claim 29, wherein

the device further comprises:

a transient class determining unit, configured to divide a full-frequency band time-domain signal of the current frame into N sub-frames, and when energy of one sub-frame is greater than a predetermined number of times of energy of a previous sub-frame of the sub-frame, determine that the signal class of the high-frequency band signal of the current frame is a transient class.