PyTorch Implementation of CenterNet(Object as Points)
- You don't need to bulid some cpp code to use Deformable Convolution used in CenterNet.
| Repo | Backbone | 0.5 mAP | Trained model |
|---|---|---|---|
| This Repo | ResNet-18 | 78.1 | |
| xingyizhou/CenterNet | ResNet-18 | 75.6 | |
| Ximilar-com/xcenternet | ResNet-18 | 70.0 | |
| xuannianz/keras-CenterNet | ResNet-50 | 72.9 | |
| bubbliiiing/centernet-keras | ResNet-50 | 77.1 |
| Repo | Backbone | mAP | Trained model |
|---|---|---|---|
| This Repo | ResNet-18 | ||
| xingyizhou/CenterNet | ResNet-18 | 28.1 |
Refer to Appendix D: Experiments on PascalVOC in the paper
- Learning rate scheduler: MultiStepLR(milestones=[45, 60], gamma=0.1)
- Augmentation: HorizontalFlip, RandomScale, RandomTranslation, RandomCrop, and Color Jittering
- Kernel size of max pooling: 3
- Learning rate scheduler: CosineDecay(per iteration)
- Augmentation: HorizontalFlip, RandomScale, RandomTranslation, RandomCrop, Mosaic, Mixup(with Mosaic + 1.0 AP), and Color Jittering
- Kernel size of max pooling: 7
- Gaussian Kernel Generation Method: followed the method proposed in Training-Time-Friendly Network for Real-Time Object Detection (It’s not carefully selected. I just think that it is more reasonable than original one.)
git clone https://github.com/developer0hye/Simple-CenterNet
cd Simple-CenterNet
if (your_os == 'Window'):
scripts/download-voc0712.bat
scripts/download-coco17.bat
else:
scripts/download-voc0712.sh
scripts/download-coco17.sh
python train.py --data ./data/voc0712.yaml --step-batch-size 32 --forward-batch-size 32 --total-epoch 70
If your gpu memory is too lower to train the model, you should try to reduce forward-batch-size.
python eval.py --data ./data/voc0712.yaml --weights your_model.pth --flip
- Random Rotation Augmentation