LiteVSR: Efficient Visual Speech Recognition by Learning from Speech Representations of Unlabeled Data

VLM
Re-assign community