Investigating and Improving Counter-Stereotypical Action Relation in Text-to-Image Diffusion Models
원문 초록
Text-to-image diffusion models consistently fail at generating counter-stereotypical action relationships (e.g., "mouse chasing cat"), defaulting to frequent stereotypes even when explicitly prompted otherwise. Through systematic investigation, we discover this limitation stems from distributional biases rather than inherent model constraints. Our key insight reveals that while models fail on rare compositions when their inversions are common, they can successfully generate similar intermediate compositions (e.g., "mouse chasing boy"). To test this hypothesis, we develop a Role-Bridging Decomposition framework that leverages these intermediates to gradually teach rare relationships without architectural modifications. We introduce ActionBench, a comprehensive benchmark specifically designed to evaluate action-based relationship generation across stereotypical and counter-stereotypical configurations. Our experiments validate that intermediate compositions indeed facilitate counter-stereotypical generation, with both automatic metrics and human evaluations showing significant improvements over existing approaches. This work not only identifies fundamental biases in current text-to-image systems but demonstrates a promising direction for addressing them through compositional reasoning.
함께 살펴볼 논문
공통 태그를 근거로 연결했습니다. 실제 인용 관계나 논문의 우수성을 뜻하지 않습니다.
Enhancing Homophily-Heterophily Separation: Relation-Aware Learning in Heterogeneous Graphs
Chaining Event Spans for Temporal Relation Grounding
A MISMATCHED Benchmark for Scientific Natural Language Inference
MonkeyOCR: Document Parsing with a Structure-Recognition-Relation Triplet Paradigm
Comparative Analysis of AI Agent Architectures for Entity Relationship Classification
Towards a More Generalized Approach in Open Relation Extraction
토론
이 논문과 연결된 포럼 글이 아직 없습니다.