C²Former: Calibrated and Complementary Transformer for RGB-Infrared Object Detection

This repo is the official implementation for C²Former: Calibrated and Complementary Transformer for RGB-Infrared Object Detection. The paper has been accepted to TGRS.

News

[2024.05.09] Code released!

[2024.04.15] Our paper is accepted!

Overview

Framework

Visualization

Results

Kaist Results

Installation

Please refer to [mmrotate installation] for installation.

Getting Started

Train with a single GPU.

python tools/train.py configs/s2anet/s2anet_c2former_fpn_1x_dota_le135.py --work-dir work_dirs/C2Former

Inference

python tools/test.py configs/s2anet/s2anet_c2former_fpn_1x_dota_le135.py work_dirs/C2Former/${CHECKPOINT_FILE} --out work_dirs/C2Former/results.pkl

Citation

If this is useful for your research, please consider cite.

@article{yuan2024c,
  title={C 2 Former: Calibrated and Complementary Transformer for RGB-Infrared Object Detection},
  author={Yuan, Maoxun and Wei, Xingxing},
  journal={IEEE Transactions on Geoscience and Remote Sensing},
  year={2024},
  publisher={IEEE}
}

@inproceedings{yuan2022translation,
  title={Translation, scale and rotation: cross-modal alignment meets RGB-infrared vehicle detection},
  author={Yuan, Maoxun and Wang, Yinyan and Wei, Xingxing},
  booktitle={European Conference on Computer Vision},
  pages={509--525},
  year={2022},
  organization={Springer}
}

yuanmaoxun/C2Former

C2Former: Calibrated and Complementary Transformer for RGB-Infrared Object Detection