ACL RD-TEC 1.0 Summarization of H94-1067
Paper Title:
MICROPHONE-INDEPENDENT ROBUST SIGNAL PROCESSING USING PROBABILISTIC OPTIMUM FILTERING
MICROPHONE-INDEPENDENT ROBUST SIGNAL PROCESSING USING PROBABILISTIC OPTIMUM FILTERING
Authors: Leonardo Neumeyer and Mitchel Weintraub
Primarily assigned technology terms:
- algorithm
- atis
- automatic speech recognizer
- cepstral-mean normalization
- cepstral-mean normalization \
- continuous speech recognizer
- data collection
- database
- feature mapping
- hidden markov
- hidden markov model
- mapping algorithm
- markov model
- model adaptation
- normalization
- pof mapping
- probabilistic optimum filtering
- processing
- recognition
- recognition system
- recognition system \
- recognizer
- robust speech recognition
- signal processing
- speech recognition
- speech recognition system
- speech recognizer
- word recognition
Other assigned terms:
- acoustic models
- approach
- case
- cepstral energy
- cepstral vector
- coefficient
- compact representation
- conditional probability
- continuous speech
- correlation
- density function
- dimensionality
- distribution
- error rate
- experimental results
- feature
- feature set
- feature space
- feature vector
- frame
- gaussian distribution
- hmm model
- hypothesis
- index
- likelihood
- mapping
- mappings
- maps
- measure
- model parameters
- noise
- noisy speech
- posterior
- probabilities
- probability
- probability density
- probability density function
- probability distributions
- recognition errors
- recognition task
- run-time
- sentence
- sentences
- signal
- signal-to-noise ratio
- speech corpus
- speech recognition performance
- speech signal
- symbols
- test data
- test set
- testing speech
- theorem
- training
- training set
- transformation
- word
- word error rate
- wsj corpus
