Speech recognition with a new hybrid architecture combining neural networks and continuous HMM.

Daniel Willett, Gerhard Rigoll

Browse the full ESANN paper archive.