Using dynamic conditional random field on single-microphone speech separation.
Yu Ting YeungTan LeeCheung-Chi LeungPublished in: ICASSP (2013)
Keyphrases
- conditional random fields
- automatic speech recognition
- hidden markov models
- crf model
- graphical models
- image labeling
- probabilistic model
- long range
- generative model
- maximum entropy
- random fields
- information extraction
- speech recognition
- sequence classification
- higher order
- speaker diarization
- max margin
- semantic segmentation
- undirected graphical models
- efficient inference
- sound source
- audio visual
- contextual features
- ordinal scale
- structured prediction
- speech signal
- approximate inference
- computer vision
- contextual information
- maximum likelihood
- image segmentation