AI-Generated Video Detection via Perceptual Straightening

TL;DR

This paper introduces ReStraV, a novel method that leverages perceptual straightening in neural representations to effectively distinguish AI-generated videos from real ones, achieving state-of-the-art detection accuracy.

Contribution

The paper proposes a new detection approach based on analyzing geometric properties of neural representations, specifically curvature and distance, for identifying AI-generated videos.

Findings

Achieves 97.17% accuracy on VidProM benchmark.

Outperforms existing detection methods significantly.

Provides insights into neural representation geometry for detection.

Abstract

The rapid advancement of generative AI enables highly realistic synthetic videos, posing significant challenges for content authentication and raising urgent concerns about misuse. Existing detection methods often struggle with generalization and capturing subtle temporal inconsistencies. We propose ReStraV(Representation Straightening Video), a novel approach to distinguish natural from AI-generated videos. Inspired by the "perceptual straightening" hypothesis -- which suggests real-world video trajectories become more straight in neural representation domain -- we analyze deviations from this expected geometric property. Using a pre-trained self-supervised vision transformer (DINOv2), we quantify the temporal curvature and stepwise distance in the model's representation domain. We aggregate statistics of these measures for each video and train a classifier. Our analysis shows that AI-generated videos exhibit significantly different curvature and distance patterns compared to real videos. A lightweight classifier achieves state-of-the-art detection performance (e.g., 97.17% accuracy and 98.63% AUROC on the VidProM benchmark), substantially outperforming existing image- and video-based methods. ReStraV is computationally efficient, it is offering a low-cost and effective detection solution. This work provides new insights into using neural representation geometry for AI-generated video detection.