OpenSR: Open-Modality Speech Recognition via Maintaining Multi-Modality Alignment.
Xize ChengTao JinLinjun LiWang LinXinyu DuanZhou ZhaoPublished in: ACL (1) (2023)
Keyphrases
- speech recognition
- multi modality
- multi modal
- medical images
- single modality
- hidden markov models
- language model
- speech synthesis
- speech processing
- pattern recognition
- information theoretic
- speech recognizer
- automatic speech recognition
- speech signal
- speech recognition systems
- mutual information
- speaker independent
- isolated word
- speaker identification
- imaging modalities
- image registration
- anatomical structures
- speech recognition technology
- multiple modalities
- image analysis
- image processing
- speech recognizers
- computer vision