AdelaiDet is an open source toolbox for multiple instance-level recognition tasks on top of Detectron2. All instance-level recognition works from our group are open-sourced here.
To date, AdelaiDet implements the following algorithms:
- FCOS
- BlendMask
- ABCNet to be released (demo)
- SOLO to be released (mmdet version)
- SOLOv2 to be released (mmdet version)
- DirectPose to be released
- CondInst to be released
All of our trained models are available in the Model Zoo.
COCO Object Detecton Baselines with FCOS
| Name | inf. time | box AP | download |
|---|---|---|---|
| FCOS_R_50_1x | 16 FPS | 38.7 | model |
| FCOS_MS_R_50_2x | 16 FPS | 41.0 | model |
| FCOS_MS_R_101_2x | 12 FPS | 43.1 | model |
| FCOS_MS_X_101_32x8d_2x | 6.6 FPS | 43.9 | model |
| FCOS_MS_X_101_64x4d_2x | 6.1 FPS | 44.7 | model |
| FCOS_MS_X_101_32x8d_dcnv2_2x | 4.6 FPS | 46.6 | model |
Except for FCOS_R_50_1x, all other models are trained with multi-scale data augmentation.
| Name | inf. time | box AP | download |
|---|---|---|---|
| FCOS_RT_MS_DLA_34_4x_shtw | 52 FPS | 39.1 | model |
| FCOS_RT_MS_DLA_34_4x | 46 FPS | 40.3 | model |
| FCOS_RT_MS_R_50_4x | 38 FPS | 40.2 | model |
If you prefer BN in FCOS heads, please try the following models.
| Name | inf. time | box AP | download |
|---|---|---|---|
| FCOS_RT_MS_DLA_34_4x_shtw_bn | 52 FPS | 38.9 | model |
| FCOS_RT_MS_DLA_34_4x_bn | 48 FPS | 39.4 | model |
| FCOS_RT_MS_R_50_4x_bn | 40 FPS | 39.3 | model |
Inference time is measured on a NVIDIA 1080Ti with batch size 1. Real-time models use shorter side 512 for inference.
COCO Instance Segmentation Baselines with BlendMask
| Model | Name | inf. time | box AP | mask AP | download |
|---|---|---|---|---|---|
| Mask R-CNN | 550_R_50_3x | 16FPS | 39.1 | 35.3 | |
| BlendMask | 550_R_50_3x | 28FPS | 38.7 | 34.5 | model |
| BlendMask | RT_R_50_4x | 30FPS | 40.1 | 34.6 | model |
| BlendMask | DLA_34_4x | 32FPS | 40.9 | 35.2 | model |
| Mask R-CNN | R_50_1x | 13FPS | 38.6 | 35.2 | |
| BlendMask | R_50_1x | 14FPS | 39.9 | 35.8 | model |
| Mask R-CNN | R_50_3x | 13FPS | 41.0 | 37.2 | |
| BlendMask | R_50_3x | 14FPS | 42.7 | 37.8 | model |
| Mask R-CNN | R_101_3x | 10FPS | 42.9 | 38.6 | |
| BlendMask | R_101_3x | 11FPS | 44.8 | 39.5 | model |
| BlendMask | R_101_dcni3_5x | 10FPS | 46.8 | 41.1 | model |
First install Detectron2 following the official guide: INSTALL.md. Then build AdelaiDet with:
git clone https://github.com/aim-uofa/AdelaiDet.git
cd AdelaiDet
python setup.py build develop
- Pick a model and its config file, for example,
fcos_R_50_1x.yaml. - Download the model
wget https://cloudstor.aarnet.edu.au/plus/s/glqFc13cCoEyHYy/download -O fcos_R_50_1x.pth - Run the demo with
python demo/demo.py \
--config-file configs/FCOS-Detection/R_50_1x.yaml \
--input input1.jpg input2.jpg \
--opts MODEL.WEIGHTS fcos_R_50_1x.pth
To train a model with "train_net.py", first setup the corresponding datasets following datasets/README.md, then run:
OMP_NUM_THREADS=1 python tools/train_net.py \
--config-file configs/FCOS-Detection/R_50_1x.yaml \
--num-gpus 8 \
OUTPUT_DIR training_dir/fcos_R_50_1x
To evaluate the model after training, run:
OMP_NUM_THREADS=1 python tools/train_net.py \
--config-file configs/FCOS-Detection/R_50_1x.yaml \
--eval-only \
--num-gpus 8 \
OUTPUT_DIR training_dir/fcos_R_50_1x \
MODEL.WEIGHTS training_dir/fcos_R_50_1x/model_final.pth
- The configs are made for 8-GPU training. To train on another number of GPUs, change the
--num-gpus. - If you want to measure the inference time, please change
--num-gpusto 1. - We set
OMP_NUM_THREADS=1by default, which achieves the best speed on our machines, please change it as needed.
If you use this toolbox in your research or wish to refer to the baseline results, please use the following BibTeX entries.
@inproceedings{tian2019fcos,
title = {{FCOS}: Fully Convolutional One-Stage Object Detection},
author = {Tian, Zhi and Shen, Chunhua and Chen, Hao and He, Tong},
booktitle = {Proc. Int. Conf. Computer Vision (ICCV)},
year = {2019}
}
@inproceedings{chen2020blendmask,
title = {{BlendMask}: Top-Down Meets Bottom-Up for Instance Segmentation},
author = {Chen, Hao and Sun, Kunyang and Tian, Zhi and Shen, Chunhua and Huang, Yongming and Yan, Youliang},
booktitle = {Proc. IEEE Conf. Computer Vision and Pattern Recognition (CVPR)},
year = {2020}
}
@inproceedings{liu2020abcnet,
title = {{ABCNet}: Real-time Scene Text Spotting with Adaptive Bezier-Curve Network},
author = {Liu, Yuliang and Chen, Hao and Shen, Chunhua and He, Tong and Jin, Lianwen and Wang, Liangwei},
booktitle = {Proc. IEEE Conf. Computer Vision and Pattern Recognition (CVPR)},
year = {2020}
}
@article{wang2019solo,
title = {{SOLO}: Segmenting Objects by Locations},
author = {Wang, Xinlong and Kong, Tao and Shen, Chunhua and Jiang, Yuning and Li, Lei},
journal = {arXiv preprint arXiv:1912.04488},
year = {2019}
}
@article{wang2020solov2,
title = {{SOLOv2}: Dynamic, Faster and Stronger},
author = {Wang, Xinlong and Zhang, Rufeng and Kong, Tao and Li, Lei and Shen, Chunhua},
journal = {arXiv preprint arXiv:2003.10152},
year = {2020}
}
@article{tian2019directpose,
title = {{DirectPose}: Direct End-to-End Multi-Person Pose Estimation},
author = {Tian, Zhi and Chen, Hao and Shen, Chunhua},
journal = {arXiv preprint arXiv:1911.07451},
year = {2019}
}
@article{tian2020conditional,
title = {Conditional Convolutions for Instance Segmentation},
author = {Tian, Zhi and Shen, Chunhua and Chen, Hao},
journal = {arXiv preprint arXiv:2003.05664},
year = {2020}
}For academic use, this project is licensed under the 2-clause BSD License - see the LICENSE file for details. For commercial use, please contact Chunhua Shen.