Joint optimization of LCMV beamforming and acoustic echo cancellation for automatic speech recognition.
Wolfgang HerbordtSatoshi NakamuraWalter KellermannPublished in: ICASSP (3) (2005)
Keyphrases
- joint optimization
- automatic speech recognition
- speech signal
- acoustic features
- sound source
- speech sounds
- speech recognizers
- speech recognition
- speech segments
- speech enhancement
- acoustic models
- hidden markov models
- word error rate
- speech recognition systems
- conversational speech
- noisy environments
- vocal tract
- speech corpus
- sparse representation
- broadcast news
- speech retrieval
- recognition errors
- frequency domain
- non stationary
- spoken words
- speaker independent
- word recognition
- speaker identification
- neural network
- speech recognizer
- linear prediction
- language model
- data points
- computer vision