paper-with-me

Papers

Uni-AdaVD: Universal Concept Erasure for Visual Generation via Orthogonal Value Decomposition

2026-07-16 · Qifan Zhou, Yuan Wang, Yanbin Hao, Xiang Wang, Kuien Liu, Richang Hong, Meng Wang arxiv

Visual generative models inevitably absorb undesirable concepts from uncurated pretraining data, making concept erasure essential for safe deployment. Existing erasure methods, however, are often architecture-specific and struggle to remove target concepts while preserving non-target content and generative priors. We present Uni-AdaVD, a universal inference-time concept erasure framework for visual generation. Uni-AdaVD treats the value space of multimodal attention as a unified intervention space and introduces encoder-aware target representation construction to localize target semantics across heterogeneous text encoders. It further combines orthogonal value decomposition with an adaptive erasing shift to suppress target semantic directions without updating the original model weights. Extensive experiments on U-Net-, DiT-, and autoregressive image generators, as well as text-to-video models, demonstrate strong performance on single- and multi-concept erasure while preserving non-target priors. These results suggest that Uni-AdaVD provides an efficient and adaptable safety mechanism for modern visual generative models. Our code is available at https://github.com/QifanZhou/Uni-AdaVD.

📄 PDF Abstract BibTeX arXiv:2607.14521

Code (1)

QifanZhou/Uni-AdaVD ★ 1

Similar Papers 제목 키워드 기반

Precise, Fast, and Low-cost Concept Erasure in Value Space: Orthogonal Complement Matters

2024-12-09 · CVPR 2025 1 · YuAn Wang, Ouxiang Li, Tingting Mu, Yanbin Hao 외

Recent success of text-to-image (T2I) generation and its increasing practical applications, enabled by diffusion models, require urgent consideration of erasing unwanted concepts, e.g., copyrighted, offensive, and unsafe…

Image GenerationNavigateText-to-Image Generation

Closing the Safety Gap: Surgical Concept Erasure in Visual Autoregressive Models

2025-09-26 · Xinhao Zhong, Yimin Zhou, Zhiqi Zhang, Junhao Li 외 arxiv

The rapid progress of visual autoregressive (VAR) models has brought new opportunities for text-to-image generation, but also heightened safety concerns. Existing concept erasure techniques, primarily designed for diffus…

Text-to-Image Generation

Bi-Erasing: A Bidirectional Framework for Concept Removal in Diffusion Models

2025-12-15 · Hao Chen, Yiwei Wang, Songze Li arxiv

Concept erasure, which fine-tunes diffusion models to remove undesired or harmful visual concepts, has become a mainstream approach to mitigating unsafe or illegal image generation in text-to-image models.However, existi…

Image Generation

EAR: Erasing Concepts from Unified Autoregressive Models

2025-06-25 · Haipeng Fan, Shiyuan Zhang, Baohunesitu, Zihang Guo 외

Autoregressive (AR) models have achieved unified and strong performance across both visual understanding and image generation tasks. However, removing undesired concepts from AR models while maintaining overall generatio…

Image Generation

Obliviate: Erasing Concepts from Autoregressive Image Generation Models

2026-06-26 · Hossein Shakibania, Jonas Henry Grebe, Tobias Braun, Ege Aktemur 외 arxiv

The widespread adoption of generative AI models has intensified concerns about misuse, including the creation of unsafe or disturbing imagery. To mitigate such issues, several concept erasure approaches have been propose…

Image Generation