Pith. sign in

REVIEW 3 cited by

RealEra: Semantic-level Concept Erasure via Neighbor-Concept Mining

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2410.09140 v1 pith:BJKHORC6 submitted 2024-10-11 cs.CV

classification cs.CV
keywords conceptsconcepterasurerealeraerasinggenerationirrelevantspecificity
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

The remarkable development of text-to-image generation models has raised notable security concerns, such as the infringement of portrait rights and the generation of inappropriate content. Concept erasure has been proposed to remove the model's knowledge about protected and inappropriate concepts. Although many methods have tried to balance the efficacy (erasing target concepts) and specificity (retaining irrelevant concepts), they can still generate abundant erasure concepts under the steering of semantically related inputs. In this work, we propose RealEra to address this "concept residue" issue. Specifically, we first introduce the mechanism of neighbor-concept mining, digging out the associated concepts by adding random perturbation into the embedding of erasure concept, thus expanding the erasing range and eliminating the generations even through associated concept inputs. Furthermore, to mitigate the negative impact on the generation of irrelevant concepts caused by the expansion of erasure scope, RealEra preserves the specificity through the beyond-concept regularization. This makes irrelevant concepts maintain their corresponding spatial position, thereby preserving their normal generation performance. We also employ the closed-form solution to optimize weights of U-Net for the cross-attention alignment, as well as the prediction noise alignment with the LoRA module. Extensive experiments on multiple benchmarks demonstrate that RealEra outperforms previous concept erasing methods in terms of superior erasing efficacy, specificity, and generality. More details are available on our project page https://realerasing.github.io/RealEra/ .

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 3 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Uncertainty Quantification for Incomplete Multi-View Data Using Divergence Measures

    cs.CV 2025-07 reject novelty 4.0 of 10

    KPHD-Net replaces KL divergence with Proper Hölder Divergence in evidential multi-view learning, adding Kalman-filtered Dempster-Shafer fusion to quantify uncertainty in classification and clustering.

  2. Robust Brain Tumor Segmentation with Incomplete MRI Modalities Using H\"older Divergence and Mutual Information-Enhanced Knowledge Transfer

    cs.CV 2025-07 reject novelty 4.0 of 10

    A parallel 3D U-Net with Hölder-divergence and mutual-information losses claims state-of-the-art missing-modality brain-tumor segmentation on BraTS 2018/2020, but the comparison is cross-paper and the theoretical proo...

  3. Erasing Concepts, Steering Generations: A Comprehensive Survey of Concept Suppression

    cs.CV 2025-05 conditional novelty 4.0 of 10

    This survey classifies concept erasure methods for text-to-image diffusion models along intervention level, optimization strategy, and semantic scope, and reviews the datasets, metrics, and benchmarks used to evaluate them.

Pith tools