Home

Awesome

Probabilistic two-stage detection

Two-stage object detectors that use class-agnostic one-stage detectors as the proposal network.

<p align="center"> <img src='docs/centernet2_teaser.jpg' align="center" height="150px"> </p>

Probabilistic two-stage detection,
Xingyi Zhou, Vladlen Koltun, Philipp Krähenbühl,
arXiv technical report (arXiv 2103.07461)

Contact: zhouxy@cs.utexas.edu. Any questions or discussions are welcomed!

Summary

Main results

All models are trained with multi-scale training, and tested with a single scale. The FPS is tested on a Titan RTX GPU. More models and details can be found in the MODEL_ZOO.

COCO

ModelCOCO val mAPFPS
CenterNet-S4_DLA_8x42.571
CenterNet2_R50_1x42.924
CenterNet2_X101-DCN_2x49.98
CenterNet2_R2-101-DCN-BiFPN_4x+4x_1560_ST56.15
CenterNet2_DLA-BiFPN-P5_24x_ST49.238

LVIS

Modelval mAP box
CenterNet2_R50_1x26.5
CenterNet2_FedLoss_R50_1x28.3

Objects365

Modelval mAP
CenterNet2_R50_1x22.6

Installation

Our project is developed on detectron2. Please follow the official detectron2 installation.

We use the default detectron2 demo script. To run inference on an image folder using our pre-trained model, run

python demo.py --config-file configs/CenterNet2_R50_1x.yaml --input path/to/image/ --opts MODEL.WEIGHTS models/CenterNet2_R50_1x.pth

Benchmark evaluation and training

Please check detectron2 GETTING_STARTED.md for running evaluation and training. Our config files are under configs and the pre-trained models are in the MODEL_ZOO.

License

Our code is under Apache 2.0 license. centernet/modeling/backbone/bifpn_fcos.py are from AdelaiDet, which follows the original non-commercial license.

Citation

If you find this project useful for your research, please use the following BibTeX entry.

@inproceedings{zhou2021probablistic,
  title={Probabilistic two-stage detection},
  author={Zhou, Xingyi and Koltun, Vladlen and Kr{\"a}henb{\"u}hl, Philipp},
  booktitle={arXiv preprint arXiv:2103.07461},
  year={2021}
}