A DNN parameter mask for the binaural reverberant speech segregation.
Yi JiangWei LiYuanyuan ZuRunsheng LiuChao MaPublished in: CISP-BMEI (2016)
Keyphrases
- speech signal
- sound source
- speech recognition
- transfer function
- computational auditory scene analysis
- automatic speech recognition
- vocal tract
- noisy environments
- hidden markov models
- speaker recognition
- speaker identification
- non stationary
- parameter values
- audio visual
- automatic speech recognition systems
- endpoint detection
- fundamental frequency
- spoken language
- speech synthesis
- machine learning
- recognition engine
- spontaneous speech
- source localization
- information retrieval
- speech processing
- genetic algorithm
- data sets
- parameter space
- parameter settings