Reproduce “Child-Oriented AIGC Video Risk Reviewing: A Benchmark and Knowledge-Supported Iterative Reasoning Framework” 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
Lewen Mi, Manyi Li, Yuling Sun, Yufan Zhang, Yuxin Shi, Yulong Bian, Xiangxian Li, Juan Liu
QVRS-E reviews potential risks to children in AI-generated videos using CAVSR, a benchmark of 605 curated videos with 26 fine-grained labels. It samples 64 frames and iteratively asks visually checkable questions, grounds answers in evidence, and reviews whether coverage is sufficient. On the 26-label task, it reaches Micro-F1 0.4498 versus 0.2414 for the same Qwen3-VL-8B backbone alone; results are benchmark-specific.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.