Pacific Communication Sciences, Inc.

Audiocodes Ltd.

Harprit S Chhatwal

Hui Li

Andrew Linkens

Mark Smith

A method of determining noise in an audio stream, the method comprising: acquiring a plurality of consecutive time frames of the audio stream each comprising samples of the audio stream; generating a discrete frequency spectrum for each frame responsive to the frame samples; partitioning the frequency spectrum of each frame into a plurality of same frequency bands; determining an audio energy for each frequency band in each frame; and determining an estimate of noise energy for each frequency band in a temporally last time frame responsive to a relatively small number of smallest values for the audio energy in the frequency band of the plurality of time frames.

Noise suppressor

Methods and apparatus determining codevectors in response to a speech signal include a first codebook member which stochastically determines the characteristics of a bi-pulse codevector representative of a target vector associated with the speech signal and for removing the bi-pulse codevector from the target vector thereby forming an intermediate target vector. A second codebook member stochastically determines the characteristics of a second bi-pulse codevector in response to the intermediate target vector. In one embodiment of the invention a third codebook member adaptively determines a first codeword in response to the target signal and forms another intermediate target signal and a fourth codebook member stochastically determines a second codeword in response to the intermediate target signal. Synthesized speech signals are determined from the first and second codevectors and from the first and second codewords formed from either the first and second codebook pairing or from the third and fourth codebook pairing. A comparator determines and chooses the synthesized speech signal having the least difference with the speech signal. The codewords or codevectors associated with the chosen synthesized speech signal are selected for transmission. In most cases, the speech signal is divided into frames and each frame is divided into subframes. In such situations, codebook searches may be determined for each subframe. In a further embodiment, an additional codebook search, a single pulse codebook search, is performed over a plurality of subframes, preferably two. In another embodiment, remainder signal are formed by removing the codewords and codevectors from the speech signal. In such an embodiment, a weighting filter is provided for weighting predetermined portions of the remainder signals prior to determining which remainder signal is representative of the synthesized speech signal having the least difference with the original speech signal.

Adaptive speech coder having code excited linear predictor with multiple

Methods and apparatus for speech coding are disclosed for converting analog speech signals to digital speech signals for transmission. The speech coder, utilizing CELP techniques, includes a first filter for filtering out the spectral information from the speech signal. The spectral information is provided for transmission. A second filter is provided for filtering out the pitch information from the speech signal and such pitch information is also provided for transmission. A codevector generator determines, in one embodiment, the characteristics of a bi-pulse codevector representative of the speech signal. In this embodiment the impulse response of the first filter is truncated for determining the codevector characteristics. In this embodiment it is also preferred to determine the codevector characteristics by conducting a numerator only search in relation to a traditional fraction used for determining codevectors. In another embodiment, the codevector generator includes a transformer for transforming codevector possibilities from being representative of pulse-like sound to being representative of noise-like sound. It is especially preferred for the transform to be a Hadamard transform. It is also preferred to scramble the transformed codevector to modify the sequency properties. In still another embodiment the bi-pulse codevector generator and the scrambled codevector generator are combined with a single pulse codevector generator. In such an embodiment, it is preferred to include a comparator for evaluating the characteristics determined by the three codebook generators and choosing the output of the one providing the best codebook vector.

Growing community of inventors

Heston, United Kingdom

Harprit S Chhatwal