Compensation for phonetic nuisance variability in speaker recognition using DNNs.
Themos StafylakisPatrick KennyVishwa GuptaMd. Jahangir AlamMarcel KockmannPublished in: Odyssey (2016)
Keyphrases
- speech recognition
- speaker independent
- automatic speech recognition
- speaker dependent
- speech sounds
- hidden markov models
- speaker verification
- language model
- speaker adaptation
- acoustic models
- real time
- speech signal
- speech recognizer
- software product line
- speaker identification
- speaker recognition
- audio visual
- pattern recognition
- speaker diarization
- intra class
- shape variations
- spoken document retrieval
- document images
- maximum likelihood
- feature set
- social networks