Combining standard and throat microphones for robust speech recognition

We present a method to combine the standard and throat microphone signals for robust speech recognition in noisy environments. Our approach is to use the probabilistic optimum filter (POF) mapping algorithm to estimate the standard microphone clean-speech feature vectors, used by standard speech recognizers, from both microphonesʹ noisy-speech feature vectors. A small untranscribed "stereo" database (noisy and clean simultaneous recordings) is required to train the POF mappings. In continuous-speech recognition experiments using SRI Internationalʹs DECIPHER recognition system, both using artificially added noise and using recorded noisy speech, the combinedmicrophone approach significantly outperforms the single-microphone approach.