Evolutionary Trigger Set Generation for DNN Black-Box Watermarking
This work addresses the need for model owners to prove ownership and combat piracy in commercial deep learning, representing an incremental improvement over prior watermarking methods.
The paper tackled the problem of intellectual property protection for deep neural networks by proposing an evolutionary algorithm to generate trigger patterns for black-box watermarking, resulting in a significant reduction in false positive rates while maintaining robustness against attacks.
The commercialization of deep learning creates a compelling need for intellectual property (IP) protection. Deep neural network (DNN) watermarking has been proposed as a promising tool to help model owners prove ownership and fight piracy. A popular approach of watermarking is to train a DNN to recognize images with certain \textit{trigger} patterns. In this paper, we propose a novel evolutionary algorithm-based method to generate and optimize trigger patterns. Our method brings a siginificant reduction in false positive rates, leading to compelling proof of ownership. At the same time, it maintains the robustness of the watermark against attacks. We compare our method with the prior art and demonstrate its effectiveness on popular models and datasets.