SphereVideo: Prototype-anchored Hyperspherical Boundary for Continual AI-generated Video Detection

2026-08-02

AI Summary

This work addresses the limited generalization and susceptibility to catastrophic forgetting in existing AI-generated video detection methods when confronted with continuously evolving generative models. To overcome these challenges, the authors propose SphereVideo, a novel framework that introduces, for the first time in this domain, a prototype-anchored hyperspherical boundary mechanism. This approach constructs a compact feature distribution centered around prototypes of real videos while explicitly repelling generated samples, thereby establishing a stable decision boundary. Additionally, SphereVideo incorporates multi-granular temporal modeling at both frame and clip levels to better capture the dynamic characteristics of authentic videos. On comprehensive evaluation benchmarks, SphereVideo achieves state-of-the-art performance, surpassing prior methods by 3.08% and 4.00% in detection accuracy on seen and unseen generated videos, respectively, effectively balancing model plasticity and stability.

Abstract

AI-generated video (AIGV) detection aims to distinguish real videos from AI-generated ones. In practice, detectors trained on existing data often fail to generalize to newly emerging generative models, making this task challenging. Therefore, continual learning (CL) is essential for improving the adaptability. However, CL frameworks for this task remain underexplored. To this end, we propose SphereVideo, a novel CL framework for AIGV detection built on two key observations. First, real videos exhibit a compact feature distribution. Based on this, we encourage real video features to cluster around a real prototype on a hypersphere while repelling AI-generated samples, thereby establishing a decision boundary. This prototype serves as a stable anchor for CL, regulating boundary evolution and mitigating catastrophic forgetting. Second, existing methods tend to rely solely on spatial artifacts as shortcuts. To enhance temporal modeling, we introduce a strategy that models the temporal dynamics of real data at both frame and clip levels. By strengthening real data modeling, this strategy further facilitates learning a real prototype and forming a stable decision boundary. Moreover, we construct a comprehensive and challenging benchmark. Extensive experiments demonstrate that SphereVideo achieves an improved plasticity-stability trade-off, outperforming prior methods by 3.08% on seen data and 4.00% on unseen AI-generated data.

Problem

• AI-generated video detection
• continual learning
• catastrophic forgetting
• generalization
• temporal modeling

Innovation

• continual learning
• hyperspherical embedding
• prototype anchoring
• temporal modeling
• AI-generated video detection