Matsushita Electric Corporation of America

Matsushita Electric Industrial Co., Ltd.

Panasonic Technologies, Inc.

Philippe Gelin

Jean-Claude Junqua

Patrick Nguyen

Philippe Rene Morin

The audio source is spectrally shaped by filtering in the time domain to approximate or emulate a standardized or target microphone input channel. The background level is adjusted by adding noise to the time domain signal prior to the onset of speech to set a predetermined background noise level based on a predetermined target. The audio source is then monitored in real time and the signal-to-noise ratio is adjusted by adding noise to the time domain signal, in real time, to maintain a signal-to-noise ratio based on a predetermined target value. The normalized audio signal may be applied to both training speech and test speech. The resultant normalization minimizes the mismatch between training and testing and also improves other speech processing functions, such as speech endpoint detection.

Robust preprocessing signal equalization system and method for normalizing to a target environment

An accurate and reliable method is provided for detecting speech from an input speech signal. A probabilistic approach is used to classify each frame of the speech signal as speech or non-speech. The speech detection method is based on a frequency spectrum extracted from each frame, such that the value for each frequency band is considered to be a random variable and each frame is considered to be an occurrence of these random variables. Using the frequency spectrums from a non-speech part of the speech signal, a known set of random variables is constructed. Next, each unknown frame is evaluated as to whether or not it belongs to this known set of random variables. To do so, a unique random variable (preferably a chi-square value) is formed from the set of random variables associated with the unknown frame. The unique variable is normalized with respect the known set of random variables and then classified as either speech or non-speech using the &ldquo;Test of Hypothesis&rdquo;. Thus, each frame that belongs to the known set of random variables is classified as non-speech and each frame that does not belong to the known set of random variables is classified as speech.

Speech detection using stochastic confidence measures on the frequency spectrum

Supervised adaptation speech is supplied to the recognizer and the recognizer generates the N-best transcriptions of the adaptation speech. These transcriptions include the one transcription known to be correct, based on a priori knowledge of the adaptation speech, and the remaining transcriptions known to be incorrect. The system applies weights to each transcription: a positive weight to the correct transcription and negative weights to the incorrect transcriptions. These weights have the effect of moving the incorrect transcriptions away from the correct one, rendering the recognition system more discriminative for the new speaker's speaking characteristics. Weights applied to the incorrect solutions are based on the respective likelihood scores generated by the recognizer. The sum of all weights (positive and negative) are a positive number. This ensures that the system will converge.

Supervised adaptation using corrective N-best decoding

The system performs unsupervised speech model adaptation using the recognizer to generate the N-best solutions for an input utterance. Each of these N-best solutions is tested by a reliable information extraction process. Reliable information is extracted by a weighting technique based on likelihood scores generated by the recognizer, or by a non-linear thresholding function. The system may be used in a single pass implementation or iteratively in a multi-pass implementation.

Growing community of inventors

Santa Barbara, CA, United States of America

Philippe Gelin