Lune

ICCV2025Top-tier venue

SuMa: A Subspace Mapping Approach for Robust and Effective Concept Erasure in Text-to-Image Diffusion Models

Kien Nguyen, Anh Tran, Cuong Pham

Abstract

The rapid growth of text-to-image diffusion models has raised concerns about their potential misuse in generating harmful or unauthorized contents. To address these issues, several Concept Erasure methods have been proposed. However, most of them fail to achieve both robustness, i.e., the ability to robustly remove the target concept., and effectiveness, i.e., maintaining image quality. While few recent techniques successfully achieve these goals for NSFW concepts, none could handle narrow concepts such as copyrighted characters or celebrities. Erasing these narrow concepts is critical in addressing copyright and legal concerns. However, erasing them is challenging due to their close distances to non-target neighboring concepts, requiring finer-grained manipulation. In this paper, we introduce Subspace Mapping (SuMa), a novel method specifically designed to achieve both robustness and effectiveness in easing these narrow concepts. SuMa first derives a target subspace representing the concept to be erased and then neutralizes it by mapping it to a reference subspace that minimizes the distance between the two. This mapping ensures the target concept is robustly erased while preserving image quality. We conduct extensive experiments with SuMa across four tasks: subclass erasure, celebrity erasure, artistic style erasure, and instance erasure and compare the results with current state-of-the-art methods. Our method achieves image quality comparable to approaches focused on effectiveness, while also yielding results that are on par with methods targeting completeness.

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

4 top-tier papers have cited it since. Know what they found before your reviewers do.

Questions to start from

Lune can draw on

the full paper · the top-tier papers that cite it · related work

Get the answer

Free to start. No credit card required.

Lineage

  • GenErase: Generalizable and Semantically-Aware Concept Erasure in Diffusion Models cites High-Resolution Image Synthesis with Latent Diffusion Models
  • Anti-I2V: Safeguarding your Photos from Malicious Image-to-video Generation cites High-Resolution Image Synthesis with Latent Diffusion Models
  • InverFill: One-Step Inversion for Enhanced Few-Step Diffusion Inpainting cites High-Resolution Image Synthesis with Latent Diffusion Models
  • InverFill: One-Step Inversion for Enhanced Few-Step Diffusion Inpainting cites Anti-I2V: Safeguarding your Photos from Malicious Image-to-video Generation
  • Visual Instruction Tuning cites High-Resolution Image Synthesis with Latent Diffusion Models
  • Visual Instruction Tuning cites Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding
  • Orthogonal Concept Erasure for Diffusion Models cites High-Resolution Image Synthesis with Latent Diffusion Models
  • Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding cites High-Resolution Image Synthesis with Latent Diffusion Models
Builds onCited byLeft to right by year; the more cited, the higher.

Cited by top-tier papers4

  1. Anti-I2V: Safeguarding your Photos from Malicious Image-to-video GenerationDuc Vu, Anh Nguyen, Chi Tran, Anh TranCVPR 2026 · 7 citations
  2. InverFill: One-Step Inversion for Enhanced Few-Step Diffusion InpaintingDuc Vu, Kien Nguyen, Trong-Tung Nguyen, Ngan Nguyen et al.CVPR 2026 · 4 citations
  3. GenErase: Generalizable and Semantically-Aware Concept Erasure in Diffusion ModelsKorada Sri Vardhana, Soma BiswasCVPR 2026
  4. Orthogonal Concept Erasure for Diffusion ModelsYuhao Sun, Lingyun Yu, Hao-Xiang Xu, Fengyuan Miao et al.ICML 2026

Builds on19

  1. An Image is Worth 16x16 Words: Transformers for Image Recognition at ScaleAlexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn et al.ICLR 2021 · 21,477 citations
  2. LoRA: Low-Rank Adaptation of Large Language ModelsEdward J. Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu et al.ICLR 2022 · 18,833 citations
  3. High-Resolution Image Synthesis with Latent Diffusion ModelsRobin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser et al.CVPR 2022 · 13,123 citations
  4. Visual Instruction TuningHaotian Liu, Chunyuan Li, Qingyang Wu, Yong Jae LeeNeurIPS 2023 · 11,349 citations
  5. Photorealistic Text-to-Image Diffusion Models with Deep Language UnderstandingChitwan Saharia, William Chan, Saurabh Saxena, Lala Li et al.NeurIPS 2022 · 8,965 citations

Related papers

  1. Prototype-Guided Concept Erasure in Diffusion ModelsYuze Cai, Jiahao Lu, Hongxiang Shi, Yichao Zhou et al.CVPR 2026 · 3 citations
  2. Concept Erasure via Attention RedirectionAmit Schechter, Rinon Gal, Ofir Kedem, Gal Chechik, Daniel Cohen-OrCVPR Findings 2026
  3. EMMA: Concept Erasure Benchmark with Comprehensive Semantic Metrics and Diverse CategoriesLu Wei, Yuta Nakashima, Noa GarciaCVPR 2026 · 4 citations
  4. MACE: Mass Concept Erasure in Diffusion ModelsShilin Lu, Zilan Wang, Leyang Li, Yanzhu Liu et al.CVPR 2024 · 40 citations
  5. ICE: Intercede Concept Erasure in Text-to-Image Diffusion ModelsYizhou Lin, Nisha Huang, Kaer Huang, Henglin Liu et al.ACM MM 2025 · 1 citation
Dusk over the sea between two cliffs drawn in fine vertical lines