Deformable TDNN with adaptive receptive fields for speech recognition.
Keyu AnYi ZhangZhijian OuPublished in: CoRR (2021)
Keyphrases
- speech recognition
- receptive fields
- biologically inspired
- natural images
- language model
- pattern recognition
- hidden markov models
- automatic speech recognition
- speech recognizer
- speech synthesis
- image representation
- speaker identification
- speech signal
- hidden layer
- speech recognition systems
- saliency map
- input space
- space time
- visual information
- semi supervised
- multiscale
- bayesian networks
- feature extraction
- computer vision
- information retrieval