Reproduce “VideoScout: Learning Agentic Active Exploration with Adaptive Reasoning Pacing for Long Video Understanding” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Weixin Xu, Zhenyu Yang, Bing Wang, Shengsheng Qian, Changsheng Xu
VideoScout addresses long-video question answering with Sequential Evidence Acquisition: it moves forward through chunks, adjusts viewing speed, carries compact notes, and locally rewatches uncertain recent footage. Two-stage training teaches this exploration policy. Across four long-video scores, it reaches an average of 48.0, versus 33.1 for an untrained Qwen2.5-VL-7B variant, but per-turn input remains bounded while runtime can increase.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.