Preprint
Aug 2026
Video-DeepResearch: Towards the Next-Generation Multimodal Deepresearch Agent
Video-DR is introduced, featuring a decoupled perception-exploration pipeline with stage-wise tool unlocking that compels exhaustive cross-frame visual grounding prior to web retrieval, enabling autonomous exploration that breaks the imitation-learning ceiling.
Zhen Fang, Yu Zeng, Wen-Xuan Huang et al.
· 1 citation
· ⚡1