SyncVSR: Data-Efficient Visual Speech Recognition with End-to-End Crossmodal Audio Token Synchronization

Documenti analoghi

• Monet: Mixture of Monosemantic Experts for Transformers
  di: Park, Jungwoo, et al.
  Pubblicazione: (2024)
• ScrewSplat: An End-to-End Method for Articulated Object Recognition
  di: Kim, Seungyeon, et al.
  Pubblicazione: (2025)
• MMS-LLaMA: Efficient LLM-based Audio-Visual Speech Recognition with Minimal Multimodal Speech Tokens
  di: Yeo, Jeong Hun, et al.
  Pubblicazione: (2025)
• AV2AV: Direct Audio-Visual Speech to Audio-Visual Speech Translation with Unified Audio-Visual Speech Representation
  di: Choi, Jeongsoo, et al.
  Pubblicazione: (2023)
• AlignVSR: Audio-Visual Cross-Modal Alignment for Visual Speech Recognition
  di: Liu, Zehua, et al.
  Pubblicazione: (2024)