Title :
Context modeling in hybrid segment-based/neural network recognition systems
Author :
Verhasselt, Jan ; Martens, Jean-Pierre
Author_Institution :
ELIS, Ghent Univ., Belgium
Abstract :
In this paper, we describe the incorporation of context-dependent models in hybrid segment-based/neural network (SB/NN) speech recognition systems. We present alternative probabilistic frameworks and evaluate them by performing speaker-independent phone recognition experiments on the TIMIT corpus. We compare their recognition performances with that of a context-independent hybrid SB/NN system and with the best published performances on this task
Keywords :
neural nets; speech recognition; TIMIT corpus; context modeling; context-dependent models; hybrid segment-based/neural network recognition systems; probabilistic frameworks; speaker-independent phone recognition experiments; Context modeling; Equations; Hidden Markov models; Intelligent networks; Multilayer perceptrons; Neural networks; Speech recognition; Writing;
Conference_Titel :
Acoustics, Speech and Signal Processing, 1998. Proceedings of the 1998 IEEE International Conference on
Conference_Location :
Seattle, WA
Print_ISBN :
0-7803-4428-6
DOI :
10.1109/ICASSP.1998.674477
