Chameleon: On the Scene Diversity and Domain Variety of AI-Generated Videos Detection

2025-03-09

AI Summary

Existing AI-generated video detection datasets suffer from severe limitations in diversity, authenticity, and complexity, hindering robust forensic analysis in open-world scenarios. To address this, we introduce Chameleon—a novel benchmark dataset that systematically integrates multiple generative models (e.g., SVD, Pika, Runway), diverse real-world source footage, dynamic camera transitions, and non-face-centric content (e.g., action and environmental generation), thereby departing from conventional face-centric detection paradigms. Chameleon leverages motion-aware modeling, cross-domain sampling, and dynamic editing techniques to construct a high-fidelity, generalizable benchmark for video forgery detection. Empirical evaluation demonstrates that models trained on Chameleon achieve significantly improved generalization across generative tools, real-world scenes, and action categories. As the first authoritative benchmark designed explicitly for complex, realistic AIGC video forensics, Chameleon establishes a new standard for evaluating detection robustness beyond controlled, face-focused settings.

Technology Category

• Computer Vision: Video Understanding & Activity Analysis
• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Fact-Checking / Misinformation Detection (NLP Focus)

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Artificial intelligence generated content (AIGC), known as DeepFakes, has emerged as a growing concern because it is being utilized as a tool for spreading disinformation. While much research exists on identifying AI-generated text and images, research on detecting AI-generated videos is limited. Existing datasets for AI-generated videos detection exhibit limitations in terms of diversity, complexity, and realism. To address these issues, this paper focuses on AI-generated videos detection and constructs a diverse dataset named Chameleon. We generate videos through multiple generation tools and various real video sources. At the same time, we preserve the videos' real-world complexity, including scene switches and dynamic perspective changes, and expand beyond face-centered detection to include human actions and environment generation. Our work bridges the gap between AI-generated dataset construction and real-world forensic needs, offering a valuable benchmark to counteract the evolving threats of AI-generated content.

Problem

• Detecting AI-generated videos with diverse scenes and domains.
• Addressing limitations in existing datasets for AI video detection.
• Expanding detection beyond face-centered to include actions and environments.

Innovation

• Constructs diverse dataset named Chameleon
• Generates videos using multiple tools and sources
• Expands detection to human actions and environments