FIRE-CIR: Fine-grained Reasoning for Composed Fashion Image Retrieval
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910118581370880 |
|---|---|
| author | Gardères, François Gauthier, Camille-Sovanneary Ponce, Jean Chen, Shizhe |
| author_facet | Gardères, François Gauthier, Camille-Sovanneary Ponce, Jean Chen, Shizhe |
| contents | Composed image retrieval (CIR) aims to retrieve a target image that depicts a reference image modified by a textual description. While recent vision-language models (VLMs) achieve promising CIR performance by embedding images and text into a shared space for retrieval, they often fail to reason about what to preserve and what to change. This limitation hinders interpretability and yields suboptimal results, particularly in fine-grained domains like fashion. In this paper, we introduce FIRE-CIR, a model that brings compositional reasoning and interpretability to fashion CIR. Instead of relying solely on embedding similarity, FIRE-CIR performs question-driven visual reasoning: it automatically generates attribute-focused visual questions derived from the modification text, and verifies the corresponding visual evidence in both reference and candidate images. To train such a reasoning system, we automatically construct a large-scale fashion-specific visual question answering dataset, containing questions requiring either single- or dual-image analysis. During retrieval, our model leverages this explicit reasoning to re-rank candidate results, filtering out images inconsistent with the intended modifications. Experimental results on the Fashion IQ benchmark show that FIRE-CIR outperforms state-of-the-art methods in retrieval accuracy. It also provides interpretable, attribute-level insights into retrieval decisions. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_09114 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | FIRE-CIR: Fine-grained Reasoning for Composed Fashion Image Retrieval Gardères, François Gauthier, Camille-Sovanneary Ponce, Jean Chen, Shizhe Computer Vision and Pattern Recognition Machine Learning Composed image retrieval (CIR) aims to retrieve a target image that depicts a reference image modified by a textual description. While recent vision-language models (VLMs) achieve promising CIR performance by embedding images and text into a shared space for retrieval, they often fail to reason about what to preserve and what to change. This limitation hinders interpretability and yields suboptimal results, particularly in fine-grained domains like fashion. In this paper, we introduce FIRE-CIR, a model that brings compositional reasoning and interpretability to fashion CIR. Instead of relying solely on embedding similarity, FIRE-CIR performs question-driven visual reasoning: it automatically generates attribute-focused visual questions derived from the modification text, and verifies the corresponding visual evidence in both reference and candidate images. To train such a reasoning system, we automatically construct a large-scale fashion-specific visual question answering dataset, containing questions requiring either single- or dual-image analysis. During retrieval, our model leverages this explicit reasoning to re-rank candidate results, filtering out images inconsistent with the intended modifications. Experimental results on the Fashion IQ benchmark show that FIRE-CIR outperforms state-of-the-art methods in retrieval accuracy. It also provides interpretable, attribute-level insights into retrieval decisions. |
| title | FIRE-CIR: Fine-grained Reasoning for Composed Fashion Image Retrieval |
| topic | Computer Vision and Pattern Recognition Machine Learning |
| url | https://arxiv.org/abs/2604.09114 |