paper-with-me

홈 › Papers

DeLeaker: Dynamic Inference-Time Reweighting For Semantic Leakage Mitigation in Text-to-Image Models

2025-10-16 · Mor Ventura, Michael Toker, Or Patashnik, Yonatan Belinkov, Roi Reichart arxiv

Text-to-Image (T2I) models have advanced rapidly, yet they remain vulnerable to semantic leakage, the unintended transfer of semantically related features between distinct entities. Existing mitigation strategies are often optimization-based or dependent on external inputs. We introduce DeLeaker, a lightweight, optimization-free inference-time approach that mitigates leakage by directly intervening on the model's attention maps. Throughout the diffusion process, DeLeaker dynamically reweights attention maps to suppress excessive cross-entity interactions while strengthening the identity of each entity. To support systematic evaluation, we introduce SLIM (Semantic Leakage in IMages), the first dataset dedicated to semantic leakage, comprising 1,130 human-verified samples spanning diverse scenarios, together with a novel automatic evaluation framework. Experiments demonstrate that DeLeaker consistently outperforms all baselines, even when they are provided with external information, achieving effective leakage mitigation without compromising fidelity or quality. These results underscore the value of attention control and pave the way for more semantically precise T2I models.

📄 PDF Abstract BibTeX arXiv:2510.15015

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Thermodynamic Focusing for Inference-Time Search: Practical Methods for Target-Conditioned Sampling and Prompted Inference

2025-12-16 · Zhan Zhang arxiv

Finding rare but useful solutions in very large candidate spaces is a recurring practical challenge across language generation, planning, and reinforcement learning. We present a practical framework, \emph{Inverted Causa…

Reinforcement Learning

Semantically-Guided Inference for Conditional Diffusion Models: Enhancing Covariate Consistency in Time Series Forecasting

2025-08-03 · Rui Ding, Hanyang Meng, Zeyang Zhang, Jielong Yang arxiv

Diffusion models have demonstrated strong performance in time series forecasting, yet often suffer from semantic misalignment between generated trajectories and conditioning covariates, especially under complex or multim…

Time Series Forecasting

DTTNet: Improving Video Shadow Detection via Dark-Aware Guidance and Tokenized Temporal Modeling

2025-11-10 · Zhicheng Li, Kunyang Sun, Rui Yao, Hancheng Zhu 외 arxiv

Video shadow detection confronts two entwined difficulties: distinguishing shadows from complex backgrounds and modeling dynamic shadow deformations under varying illumination. To address shadow-background ambiguity, we …

Shadow Detection

ARM: A Confidence-Based Adversarial Reweighting Module for Coarse Semantic Segmentation

2020-09-11 · Jingchao Liu, Ye Du, Zehua Fu, Qingjie Liu 외

Coarsely-labeled semantic segmentation annotations are easy to obtain, but therefore bear the risk of losing edge details and introducing background pixels. Impeded by the inherent noise, existing coarse annotations are …

Semantic Segmentation

ADE: Adaptive Dictionary Embeddings -- Scaling Multi-Anchor Representations to Large Language Models

2026-04-27 · Orhan Demirci, Sezer Aptourachman, Aydın Kaya arxiv

Word embeddings are fundamental to natural language processing, yet traditional approaches represent each word with a single vector, creating representational bottlenecks for polysemous words and limiting semantic expres…

Text Classification