CoLM-DSR: Leveraging Neural Codec Language Modeling for Multi-Modal Dysarthric Speech Reconstruction.
Xueyuan ChenDongchao YangDingdong WangXixin WuZhiyong WuHelen MengPublished in: CoRR (2024)
Keyphrases
- multi modal
- language modeling
- audio visual
- language model
- speech recognition
- finite state transducers
- retrieval model
- query expansion
- information retrieval
- cross lingual
- n gram
- probabilistic model
- multi modality
- high dimensional
- document retrieval
- test collection
- single modality
- text classification
- image annotation
- relevance model
- video search