Verus-Insight: Towards Explainable Artifact Detection in AI-Generated Videos
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 identif
Reading in Plain English mode — toggle below to switch
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 identif
As AI-generated content becomes indistinguishable from real, robust detection methods are critical for maintaining trust and combating misinformation. This work directly addresses the explainability and reliability of such detectors, which is vital for deploying trustworthy AI systems in sensitive applications.