431195

Discriminative Training of Acoustic Models Applied to Domains with Unreliable Transcripts

Mathias, Lambert ; Yegnanarayanan, Girija ; Fritsch, Juergen

Center for Language & Speech Process., Johns Hopkins Univ., Baltimore, MD, USA

1

2005

March 18-23, 2005

109

112

learning (artificial intelligence); maximum likelihood estimation; medical signal processing; speech recognition; ASR systems; acoustic model discriminative training; automatic speech recognition systems; frame filtering; lattice-based discriminative training; maximum likelihood discriminative training; medical report transcription; medical speech data; speaker-adapted acoustic models; speech data training transcripts; unreliable transcript domains; word error rate reduction; Automatic speech recognition; Biomedical acoustics; Error analysis; Filtering; Maximum likelihood estimation; Natural languages; Parameter estimation; Robustness; Speech processing; Training data;

English

ieee

Acoustics, Speech, and Signal Processing, 2005. Proceedings. (ICASSP '05). IEEE International Conference on

1520-6149

0-7803-8874-7

conf

10.1109/ICASSP.2005.1415062

1415062