Subband-based Speech Recognition

In the framework of Hidden Markov Models (HMM) or hybrid HMM/Artiicial Neural Network (ANN) systems, we present a new approach t o wards automatic speech recognition (ASR). The general idea is to divide up the full frequency band (represent e d i n t e r m s o f critical bands) into several subbands, compute phone probabilities for each sub-band on the basis of subband acoustic features, perform dynamic programming independently for each band, and merge the subband recognizers (recombining the respective, possibly weighted, scores) at some segmental level corresponding to temporal anchor points. The results presented in this paper connrm some preliminary tests reported earlier. On both isolated word and continuous speech tasks, it is indeed shown that even using quite simple recombination strategies, this subband ASR approach can yield at least comparable performance on clean speech while providing better robustness in the case of narrowband noise.