Speaker-Independent Speech Animation Using Perceptual Loss Functions and Synthetic Data.
Danny WebsdaleSarah TaylorBen MilnerPublished in: IEEE Trans. Multim. (2022)
Keyphrases
- synthetic data
- loss function
- speaker independent
- speech recognition
- speaker dependent
- hidden markov models
- speaker identification
- speech recognizer
- digit recognition
- speech recognition systems
- pairwise
- speech signal
- automatic speech recognition
- support vector
- acoustic models
- speaker adaptation
- real image data
- real world
- pattern recognition
- data sets
- risk minimization
- noisy environments
- low level
- reproducing kernel hilbert space
- spoken language
- broadcast news
- gaussian mixture model
- loss minimization
- speaker recognition
- language model
- information extraction
- information retrieval
- feature extraction
- convex loss functions
- similarity measure
- speaker diarization
- model selection