OSCAR: Optimization-Steered Agentic Planning for Composed Image Retrieval
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866915786244751360 |
|---|---|
| author | Wang, Teng Shan, Rong Lin, Jianghao Wu, Junjie Xu, Tianyi Zhang, Jianping Chen, Wenteng Zhang, Changwang Wang, Zhaoxiang Zhang, Weinan Wang, Jun |
| author_facet | Wang, Teng Shan, Rong Lin, Jianghao Wu, Junjie Xu, Tianyi Zhang, Jianping Chen, Wenteng Zhang, Changwang Wang, Zhaoxiang Zhang, Weinan Wang, Jun |
| contents | Composed image retrieval (CIR) requires complex reasoning over heterogeneous visual and textual constraints. Existing approaches largely fall into two paradigms: unified embedding retrieval, which suffers from single-model myopia, and heuristic agentic retrieval, which is limited by suboptimal, trial-and-error orchestration. To this end, we propose OSCAR, an optimization-steered agentic planning framework for composed image retrieval. We are the first to reformulate agentic CIR from a heuristic search process into a principled trajectory optimization problem. Instead of relying on heuristic trial-and-error exploration, OSCAR employs a novel offline-online paradigm. In the offline phase, we model CIR via atomic retrieval selection and composition as a two-stage mixed-integer programming problem, mathematically deriving optimal trajectories that maximize ground-truth coverage for training samples via rigorous boolean set operations. These trajectories are then stored in a golden library to serve as in-context demonstrations for online steering of VLM planner at online inference time. Extensive experiments on three public benchmarks and a private industrial benchmark show that OSCAR consistently outperforms SOTA baselines. Notably, it achieves superior performance using only 10% of training data, demonstrating strong generalization of planning logic rather than dataset-specific memorization. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_08603 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | OSCAR: Optimization-Steered Agentic Planning for Composed Image Retrieval Wang, Teng Shan, Rong Lin, Jianghao Wu, Junjie Xu, Tianyi Zhang, Jianping Chen, Wenteng Zhang, Changwang Wang, Zhaoxiang Zhang, Weinan Wang, Jun Artificial Intelligence Composed image retrieval (CIR) requires complex reasoning over heterogeneous visual and textual constraints. Existing approaches largely fall into two paradigms: unified embedding retrieval, which suffers from single-model myopia, and heuristic agentic retrieval, which is limited by suboptimal, trial-and-error orchestration. To this end, we propose OSCAR, an optimization-steered agentic planning framework for composed image retrieval. We are the first to reformulate agentic CIR from a heuristic search process into a principled trajectory optimization problem. Instead of relying on heuristic trial-and-error exploration, OSCAR employs a novel offline-online paradigm. In the offline phase, we model CIR via atomic retrieval selection and composition as a two-stage mixed-integer programming problem, mathematically deriving optimal trajectories that maximize ground-truth coverage for training samples via rigorous boolean set operations. These trajectories are then stored in a golden library to serve as in-context demonstrations for online steering of VLM planner at online inference time. Extensive experiments on three public benchmarks and a private industrial benchmark show that OSCAR consistently outperforms SOTA baselines. Notably, it achieves superior performance using only 10% of training data, demonstrating strong generalization of planning logic rather than dataset-specific memorization. |
| title | OSCAR: Optimization-Steered Agentic Planning for Composed Image Retrieval |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2602.08603 |