Reproduce “StreamClarify: Proactive Clarification for Grounded Streaming Video Question Answering” 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
Zhuchen Shao, Bo Chen, Klara Nahrstedt
StreamClarify lets streaming video assistants defer answers when a binary pre-answer judgment about reliance on video evidence has high entropy. It requests user guidance and can search a separate archive. Without clarification, LLaVA-OV-1.5 gains 8.2 percentage points over StreamBridge on OVO-Bench; this benchmark does not establish natural dialogue or long-video accuracy gains.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.