Leveraging Pre-Trained Visual Models for AI-Generated Video Detection

TL;DR

This paper introduces a method that uses pre-trained visual models to effectively detect AI-generated videos, achieving over 90% accuracy without extensive additional training.

Contribution

The study presents a novel approach leveraging pre-trained visual models for generic AI video detection, addressing limitations of existing deepfake-focused methods.

Findings

Achieves over 90% detection accuracy on VID-AID dataset.

Requires minimal additional training, using features from pre-trained models.

Validated on a diverse dataset of 10,000 AI-generated and 4,000 real videos.

Abstract

Recent advances in Generative AI (GenAI) have led to significant improvements in the quality of generated visual content. As AI-generated visual content becomes increasingly indistinguishable from real content, the challenge of detecting the generated content becomes critical in combating misinformation, ensuring privacy, and preventing security threats. Although there has been substantial progress in detecting AI-generated images, current methods for video detection are largely focused on deepfakes, which primarily involve human faces. However, the field of video generation has advanced beyond DeepFakes, creating an urgent need for methods capable of detecting AI-generated videos with generic content. To address this gap, we propose a novel approach that leverages pre-trained visual models to distinguish between real and generated videos. The features extracted from these pre-trained models, which have been trained on extensive real visual content, contain inherent signals that can help distinguish real from generated videos. Using these extracted features, we achieve high detection performance without requiring additional model training, and we further improve performance by training a simple linear classification layer on top of the extracted features. We validated our method on a dataset we compiled (VID-AID), which includes around 10,000 AI-generated videos produced by 9 different text-to-video models, along with 4,000 real videos, totaling over 7 hours of video content. Our evaluation shows that our approach achieves high detection accuracy, above 90% on average, underscoring its effectiveness. Upon acceptance, we plan to publicly release the code, the pre-trained models, and our dataset to support ongoing research in this critical area.