Stable Distillation: Regularizing Continued Pre-Training for Low-Resource Automatic Speech Recognition.
Ashish SethSreyan GhoshSrinivasan UmeshDinesh ManochaPublished in: ICASSP (2024)
Keyphrases
- automatic speech recognition
- speech recognition
- discriminative training
- speech signal
- hidden markov models
- spoken words
- broadcast news
- speech retrieval
- noisy environments
- word error rate
- conversational speech
- acoustic features
- acoustic models
- word recognition
- image processing
- spontaneous speech
- recognition errors
- noisy images
- language model
- multiresolution
- cl sr
- speech sounds
- pattern recognition