ECCV2026: Parallel Autoregressive Decoding for Omni-modal Video Captioning

MikeShou1 · x · 2026-07-09

This upcoming ECCV2026 paper proposes a parallel autoregressive decoding method for omni-modal Dense Video Captioning tasks. The approach aims to enhance the model's comprehension of video content and improve multimodal generation efficiency.

Related event: Parallel Autoregressive Decoding Framework Boosts Dense Video Captioning(2 posts)→

Original post →

More from Research

Research channel →