Ai video detector improves artifact spotting and explanation accuracy
Verus-Insight: Towards Explainable Artifact Detection in AI-Generated Videos
Computer Vision and Pattern Recognition
Summary
Detecting videos created by AI can be tricky because the fake parts might be subtle or hard to identify. The authors studied how detection systems can be better designed by breaking the process into three parts: deciding what to check, confirming if what is checked really shows signs of AI creation, and combining all the findings to make a final call. They built a system called Verus-Insight that learns from detailed human input to do these steps well. Their tests show that Verus-Insight not only spots AI-generated videos better but also explains its decisions more clearly by pointing out the real clues.
What this means in practice
- •For video security teams: Identify AI-generated videos reliably by detecting and explaining subtle generation artifacts in video content.
- •For social media platform operators: Improve detection of fake videos on platforms by combining artifact evidence with clear explanations for source prediction.
Authors
Zhuoran Zhang, Yang Shi, Qixun Wang, Tengfei Liu, Yuqi Tang, Xuanyu Zhu, Yizhuo Jia, Jingyun Hua, Shixuan Liu, Xianghua Ying, Haotian Wang, Yuanxing Zhang, Pengfei Wan, Wenjing Yang
Abstract
Recent advances in video generation have made synthetic videos increasingly realistic, placing greater demands on reliable AI-generated video detection. Existing artifact-based detectors typically identify generation artifacts as evidence for source prediction, yet they may either fail to recognize subtle artifacts or reach the correct source decision without faithfully identifying the evidence underlying it. In this work, we systematically investigate how to design and train effective artifact detectors and formulate artifact detection as a structured process comprising three complementary capabilities: Inspection Planning, which determines what should be examined; Evidence Verification, which verifies whether the inspected content exhibits genuine generation artifacts; and Evidence Aggregation, which integrates the verified evidence into the final source decision. Building on this formulation, we introduce Verus-Insight, trained with human-anchored full-process supervision and progressive capability training to strengthen both fine-grained artifact detection and video-level detection. We further introduce Verus-Bench, a benchmark with fine-grained human annotations of perceptible artifacts for evaluating artifact detection. Extensive experiments across diverse benchmarks demonstrate that Verus-Insight achieves strong performance in both fine-grained artifact analysis and AI-generated video detection, with substantial gains in recovering human-perceived artifacts. Further analyses reveal that the three capabilities are complementary yet distinct, and that Verus-Insight more reliably converts relevant inspection targets into supported artifact evidence without increasing spurious artifact claims.