AuthenLoRA: Entangling Stylization with Imperceptible Watermarks for Copyright-Secure LoRA Adapters

TL;DR

AuthenLoRA introduces a novel watermarking method integrated into LoRA training that ensures imperceptible, robust traceability of generated images without degrading stylization quality, addressing security and copyright concerns.

Contribution

It proposes a unified framework embedding watermarks into LoRA training, combining stylization and watermarking, with an expanded architecture and regularization to reduce false positives.

Findings

Achieves high-fidelity stylization with embedded watermarks.

Demonstrates robust watermark propagation in generated images.

Significantly lowers false-positive detection rates.

Abstract

Low-Rank Adaptation (LoRA) offers an efficient paradigm for customizing diffusion models, but its ease of redistribution raises concerns over unauthorized use and the generation of untraceable content. Existing watermarking techniques either target base models or verify LoRA modules themselves, yet they fail to propagate watermarks to generated images, leaving a critical gap in traceability. Moreover, traceability watermarking designed for base models is not tightly coupled with stylization and often introduces visual degradation or high false-positive detection rates. To address these limitations, we propose AuthenLoRA, a unified watermarking framework that embeds imperceptible, traceable watermarks directly into the LoRA training process while preserving stylization quality. AuthenLoRA employs a dual-objective optimization strategy that jointly learns the target style distribution and the watermark-induced distribution shift, ensuring that any image generated with the watermarked LoRA reliably carries the watermark. We further design an expanded LoRA architecture for enhanced multi-scale adaptation and introduce a zero-message regularization mechanism that substantially reduces false positives during watermark verification. Extensive experiments demonstrate that AuthenLoRA achieves high-fidelity stylization, robust watermark propagation, and significantly lower false-positive rates compared with existing approaches. Open-source implementation is available at: https://github.com/ShiFangming0823/AuthenLoRA