Peripheral features for HMM-based speech recognition.

Takashi Fukuda, Masashi Takigawa, Tsuneo Nitta

Browse the full ICASSP paper archive.