The source code and dataset will be publicly available soon.

Wen Dong, Haiyang Mei, Yinglian Ji, Yutong Jiang, Ziqi Wei, Shengfeng He, Xin Yang

[Paper] [Project Page]

In this work, we propose the DPNet, a novel learning-based detector that requires only a single RGB frame at the beginning of a sequence to detect high-speed objects over a 5-second duration, 25 times longer than prior methods.

• Python 3.8.20
• Pytorch 1.12.1
• Torchvision 0.13.1
• CUDA 10.2.0
• numpy 1.24.4

The dataset can be obtained here.

The pretrained model weight can be obtained here.

Run:

python train.py

Run:

python val.py

Run:

python infer.py

Please cite our paper if you find it is useful:

@article{dong2026you,
  title={You Only Look Intensity Once: Event-Driven Long-Term High-Speed Object Detection},
  author={Dong, Wen and Mei, Haiyang and Ji, Yinglian and Jiang, Yutong and Wei, Ziqi and He, Shengfeng and Yang, Xin},
  journal={International Journal of Computer Vision},
  volume={134},
  number={4},
  pages={149},
  year={2026},
  publisher={Springer}
}