Hybrid system combining expert-TDNNs and HMMs for continuous speech recognition.

Laurence Devillers, Christian Dugast

Browse the full ICASSP paper archive.