Speech gain quantization strategy
View Patent ↗A speech encoder that analyzes and classifies each frame of speech as being periodic-like speech or non-periodic like speech where the speech encoder performs a different gain quantization process depending if the speech is periodic or not. If the speech is periodic, the improved speech encoder obtains the pitch gains from the unquantized weighted speech signal and performs a pre-vector quantization of the adaptive codebook gain G P for each subframe of the frame before subframe processing begins and a closed-loop delayed decision vector quantization of the fixed codebook gain G C . If the frame of speech is non-periodic, the speech encoder may use any known method of gain quantization. The result of quantizing gains of periodic speech in this manner results in a reduction of the number of bits required to represent the quantized gain information and for periodic speech, the ability to use the quantized pitch gain for the current subframe to search the fixed codebook for the fixed codebook excitation vector for the current subframe. Alternatively, the new gain quantization process which was used only for periodic signals may be extended to non-periodic signals as well. This second strategy results in a slightly higher bit rate than that for periodic signals that use the new gain quantization strategy, but is still lower than the prior art's bit rate. Yet another alternative is to use the new gain quantization process for all speech signals without distinguishing between periodic and non-periodic signals.
1. A speech encoding system that receives an input speech signal, the speech encoding system comprising:
a frame processor for processing a frame of the input speech signal, the frame processor including:
a pitch gain generator that derives unquantized pitch gains; and
a first vector quantizer that receives the unquantized pitch gains and generates quantized pitch gains;
a subframe processor for processing a subframe; and
a periodic signal detector that determines whether the input speech signal is periodic or non-periodic;
wherein the frame processor applies a different gain quantization process when the input speech signal is determined to be periodic than when the input speech signal is determined to be non-periodic, and wherein the speech encoding system is configured to convert the input speech signal into an encoded speech using results from the frame processor and the subframe processor.
2. The speech encoding system of claim 1 , wherein the frame processor allocates less bits for coding the frame if the periodic signal detector determines that the speech signal is periodic than if the periodic signal detector determines that the speech signal is non-periodic.
3. The speech encoding system of claim 1 further comprising a filter that modifies the input speech signal into an unquantized weighted speech signal or an unquantized linear prediction coding residual speech signal.
4. The speech encoding system of claim 1 , wherein the input speech signal is an unquantized original speech signal.
5. The speech encoding system of claim 1 , wherein the input speech signal is an unquantized modified speech signal.
6. A speech encoding system that receives an input speech signal, the speech encoding system comprising:
a frame processor for processing a frame of the input speech signal, the frame processor including:
a pitch gain generator that derives unquantized pitch gains; and
a first vector quantizer that receives the unquantized pitch gains and generates quantized pitch gains;
a subframe processor for processing a subframe; and
a periodic signal detector that determines whether the input speech signal is a periodic input speech signal or a non-periodic input speech signal;
wherein the frame processor modifies the input speech signal to generate a modified input speech signal that is more periodic when the input speech signal is determined to be non-periodic and applies the same gain quantization process to the periodic input speech signal and the modified input speech signal, and wherein the speech encoding system is configured to convert the input speech signal into an encoded speech using results from the frame processor and the subframe processor.
7. The speech encoding system of claim 6 further comprising a filter that modifies the input speech signal into an unquantized weighted speech signal or an unquantized linear prediction coding residual speech signal.
8. The speech encoding system of claim 6 , wherein the input speech signal is an unquantized original speech signal.
9. The speech encoding system of claim 6 , wherein the input speech signal is an unquantized modified speech signal.
10. A method for use in a speech encoding system that receives an input speech signal, the method comprising:
processing a frame of the input speech signal using a frame processor by:
deriving unquantized pitch gains using a pitch gain generator; and
receiving the unquantized pitch gains by a first vector quantizer and generating quantized pitch gains;
subframe processing using a subframe processor;
determining whether the input speech signal is periodic or non-periodic;
applying a different gain quantization process when the input speech signal is determined to be periodic than when the input speech signal is determined to be non-periodic, and
converting the input speech signal into an encoded speech using results from the frame processor and the subframe processor.
11. The method of claim 10 further comprising allocating less bits for coding the frame if the periodic signal detector determines that the speech signal is periodic than if the periodic signal detector determines that the speech signal is non-periodic.
12. The method of claim 10 modifying the input speech signal into an unquantized weighted speech signal or an unquantized linear prediction coding residual speech signal.
13. The method of claim 10 , wherein the input speech signal is an unquantized original speech signal.
14. The method of claim 10 , wherein the input speech signal is an unquantized modified speech signal.