5 papers
Beyond Text Prompts: Precise Concept Erasure through Text-Image Collaboration
Jun Li, Lizhi Xiong, Ziqiang Li +4
Text-to-image generative models have achieved impressive fidelity and diversity, but can inadvertently produce unsafe or undesirable content due to implicit biases embedded in larg…
Scaling Exposes the Trigger: Input-Level Backdoor Detection in Text-to-Image Diffusion Models via Cross-Attention Scaling
Zida Li, Jun Li, Yuzhe Sha +3
Text-to-image (T2I) diffusion models have achieved remarkable success in image synthesis, but their reliance on large-scale data and open ecosystems introduces serious backdoor sec…
Organizing Unstructured Image Collections using Natural Language
Mingxuan Liu, Zhun Zhong, Jun Li +3
In this work, we introduce and study the novel task of Open-ended Semantic Multiple Clustering (OpenSMC). Given a large, unstructured image collection, the goal is to automatically…
Enhancing Supervised Composed Image Retrieval via Reasoning-Augmented Representation Engineering
Jun Li, Hongjian Dou, Zhenyu Zhang +3
Composed Image Retrieval (CIR) presents a significant challenge as it requires jointly understanding a reference image and a modified textual instruction to find relevant target im…
A Comprehensive Survey on Visual Concept Mining in Text-to-image Diffusion Models
Ziqiang Li, Jun Li, Lizhi Xiong +2
Text-to-image diffusion models have made significant advancements in generating high-quality, diverse images from text prompts. However, the inherent limitations of textual signals…