Integration of diverse recognition methodologies through reevaluation of N-best sentence hypotheses
HLT '91 Proceedings of the workshop on Speech and Natural Language
Toward a real-time spoken language system using commercial hardware
HLT '90 Proceedings of the workshop on Speech and Natural Language
A comparison of several approximate algorithms for finding multiple (N-best) sentence hypotheses
ICASSP '91 Proceedings of the Acoustics, Speech, and Signal Processing, 1991. ICASSP-91., 1991 International Conference
IEEE Transactions on Computers
Speech recognition using segmental neural nets
ICASSP'92 Proceedings of the 1992 IEEE international conference on Acoustics, speech and signal processing - Volume 1
| Hi-index | 0.00 |
Until recently, state-of-the-art, large-vocabulary, continuous speech recognition (CSR) has employed Hidden Markov Modeling (HMM) to model speech sounds. However, the limitations of HMMs in modeling dependency across phonetic segments have been known for some time. Last year, we presented the concept of a "Segmental Neural Net" (SNN) for phonetic modeling in continuous speech recognition (CSR) and demonstrated that a feed-forward neural network, used within a hybrid SNN/HMM system, is able to reduce by 20% the word error rate over the baseline HMM system. In this paper we describe two developments over the initial system. First, we present a novel way to generate fixed length segment representations based on the Discrete Cosine Transform (DCT). Second, we demonstrate that an Elliptical Basis Function (EBF) Network can be used in the same hybrid framework.