ACL RD-TEC 1.0 Summarization of H94-1068
Paper Title:
VIICROPHONE ARRAYS AND NEURAL NETWORKS FOR ROBUST SPEECH RECOGNITION
VIICROPHONE ARRAYS AND NEURAL NETWORKS FOR ROBUST SPEECH RECOGNITION
Authors: C. Che and Q. Lin and J. Pearson and B. de Vries and J. Flanagan
Primarily assigned technology terms:
- array microphone
- cepstrum normalization
- classification
- database
- dtw classification
- hmm-based recognizer
- identification
- learning
- line array microphone
- mean subtraction
- microphone array
- neural net
- neural network
- neural networks
- neural-network
- noise compensation
- normalization
- processing
- processor
- recognition
- recognition system
- recognizer
- recording environment
- resource management
- robust speech recognition
- sampling
- speaker identification
- speaker identification \
- speech recognition
- speech recognition system
- speech recognizer
- transfer function
- word recognition
Other assigned terms:
- approach
- array
- background noise
- close-talking speech
- codebook
- coefficient
- continuous speech
- euclidean distance
- evaluations
- experimental results
- frame
- grammars
- language models
- learning rate
- likelihood
- measure
- method
- model combination
- noise
- performance comparison
- procedure
- recognition accuracy
- resource management task
- sentences
- signal
- signal-to-noise ratio
- speech corpus
- speech database
- system design
- term
- terms
- testing set
- training
- user
- utterance
- word
- words
