VICTOR: Dataset Copyright Auditing in Video Recognition Systems

TL;DR

VICTOR introduces a novel, stealthy method for auditing copyrighted datasets in video recognition systems by subtly modifying samples to detect unauthorized use, addressing the unique challenges posed by video data's temporal dimension.

Contribution

This paper presents the first approach for dataset copyright auditing in video recognition, leveraging minimal sample modifications to reveal unauthorized dataset usage.

Findings

VICTOR effectively distinguishes between original and modified datasets across multiple models.

The method remains robust against various perturbations to videos and models.

Experiments demonstrate superior performance over existing image-based auditing solutions.

Abstract

Video recognition systems are increasingly being deployed in daily life, such as content recommendation and security monitoring. To enhance video recognition development, many institutions have released high-quality public datasets with open-source licenses for training advanced models. At the same time, these datasets are also susceptible to misuse and infringement. Dataset copyright auditing is an effective solution to identify such unauthorized use. However, existing dataset copyright solutions primarily focus on the image domain; the complex nature of video data leaves dataset copyright auditing in the video domain unexplored. Specifically, video data introduces an additional temporal dimension, which poses significant challenges to the effectiveness and stealthiness of existing methods. In this paper, we propose VICTOR, the first dataset copyright auditing approach for video recognition systems. We develop a general and stealthy sample modification strategy that enhances the output discrepancy of the target model. By modifying only a small proportion of samples (e.g., 1%), VICTOR amplifies the impact of published modified samples on the prediction behavior of the target models. Then, the difference in the model's behavior for published modified and unpublished original samples can serve as a key basis for dataset auditing. Extensive experiments on multiple models and datasets highlight the superiority of VICTOR. Finally, we show that VICTOR is robust in the presence of several perturbation mechanisms to the training videos or the target models.