A factorial HMM approach to simultaneous recognition of isolated digits spoken by multiple talkers on one audio channel.

Ameya N. Deoras, Mark Hasegawa-Johnson

Browse the full ICASSP paper archive.