On training the recurrent neural network encoder-decoder for large vocabulary end-to-end speech recognition.

Liang Lu, Xingxing Zhang, Steve Renals

Browse the full ICASSP paper archive.