---
license: cc-by-nc-4.0
language:
- en
tags:
- retrieval
- cir
- cirr
size_categories:
- 1K<n<10K
---
# CoLLM: A Large Language Model for Composed Image Retrieval

[[Project Page](https://collm-cvpr25.github.io/)] [[Code](https://github.com/hmchuong/CoLLM)]

*Refined CIRR benchmark from the paper. Please refer to [CIRR](https://github.com/lucas-ventura/CoVR?tab=readme-ov-file#cirr) to obtain the full dataset.*

**Accepted at CVPR 2025**

Work is a part of Summer Internship 2024 at [Amazon](https://www.amazon.science/)

Please refer to our [paper](https://arxiv.org/abs/2503.19910) for details.

## Citation

If you find CoLLM useful in your research, please cite the following paper:
```latex
@InProceedings{huynh2025collm,
    author    = {Huynh, Chuong and Yang, Jinyu and Tawari, Ashish and Shah, Mubarak and Tran, Son and Hamid, Raffay and Chilimbi, Trishul and Shrivastava, Abhinav},
    title     = {CoLLM: A Large Language Model for Composed Image Retrieval},
    booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
    year      = {2025}
}
```