Detecting AI-Generated Video: A Vision-Language Dual-View Survey

2026-07-12

AI Summary

This work addresses the growing challenge posed by increasingly photorealistic AI-generated videos, for which conventional detection methods relying on low-level artifacts have become inadequate. The study reframes deepfake detection as a "factual fidelity verification" task and introduces a novel vision–language dual-perspective classification framework. It establishes a four-tiered detection architecture that integrates intrinsic cue analysis, spatiotemporal consistency modeling, cross-modal reasoning, and world-knowledge verification guided by linguistic priors. Through a systematic review of 212 related works, the paper comprehensively synthesizes generative paradigms, detection methodologies, evaluation metrics, and benchmark datasets. This holistic approach advances the field beyond artifact-based matching toward evidence-driven semantic validation, offering both theoretical foundations and practical pathways for developing robust, interpretable, and trustworthy detection systems.

Abstract

The evolving realism of AI-generated Videos (AIGC-V) is rapidly rendering traditional artifact-centric detection insufficient, necessitating a paradigm shift from low-level inspection to high-level semantic verification. This paper presents a comprehensive survey of AIGC-V detection, reframing the task as Factual Fidelity Verification, which asks whether the events, entities, and physical processes depicted in a video are consistent with real-world facts. To systematize this rapidly evolving field, we propose a Vision-Language Dual-View taxonomy that organizes existing methods into a hierarchical, four-layer landscape, spanning intrinsic cue analysis, spatiotemporal consistency modeling, cross-modal consistency reasoning, and language-guided world-level reasoning. This dual-view framing highlights a fundamental transition from artifact matching in traditional deepfake detection to evidence-based semantic verification enabled by vision-language models and agentic reasoning pipelines. Based on a systematic review of 221 works, we synthesize AIGC-V generation paradigms, survey the landscape of detection methods, and review evaluation metrics and benchmarks in line with proposed views. Finally, we discuss current challenges and identify promising directions toward robust, explainable, and trustworthy detection.

Problem

• AI-Generated Video
• Factual Fidelity Verification
• Vision-Language Models
• Deepfake Detection
• Semantic Consistency

Innovation

• Vision-Language Dual-View
• Factual Fidelity Verification
• Semantic Verification
• AIGC-V Detection
• Cross-Modal Reasoning