CoLLM: A Large Language Model for Composed Image Retrieval

[Project Page] [Code]

Refined Fashion-IQ benchmark from the paper. Please refer to Fashion-IQ to obtain the full dataset.

Accepted at CVPR 2025

Work is a part of Summer Internship 2024 at Amazon

Please refer to our paper for details.

Citation

If you find CoLLM useful in your research, please cite the following paper:

@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}
}