Fashion IQ
Emerging7papers using it
2025first seen
The 'Fashion IQ' dataset is a large-scale fashion-specific visual question answering dataset designed to evaluate compositional reasoning in composed image retrieval tasks by requiring analysis of visual attributes in reference and candidate images.
Papers using Fashion IQ (7)
- FIRE-CIR: Fine-grained Reasoning for Composed Fashion Image RetrievalMCoT-MVS: Multi-level Vision Selection by Multi-modal Chain-of-Thought Reasoning for Composed Image RetrievalReCALL: Recalibrating Capability Degradation for MLLM-based Composed Image RetrievalX-Aligner: Composed Visual Retrieval without the Bells and WhistlesCIR-CoT: Towards Interpretable Composed Image Retrieval via End-to-End Chain-of-Thought ReasoningTMCIR: Token Merge Benefits Composed Image RetrievalCoLLM: A Large Language Model for Composed Image Retrieval