Yolov5 for Oriented Object Detection Improved by BDAM

The code for the implementation of “Yolov5 + [BDAM]+ Circular Smooth Label”.

Results and Models

The results on DOTAv1.5_subsize1024_gap200_rate1.0 test-dev set are shown in the table below. (password:0620)

Model ^(link)	Size ^(pixels)	TTA ^{(multi-scale/ rotate testing)}	OBB mAP^{test ^0.5 DOTAv1.5}	Speed ^{CPU b1 (ms)}	Speed ^{2080Ti b1 (ms)}	Speed ^{2080Ti b16 (ms)}	params ^(M)	FLOPs ^{@1024 (B)}
yolov5x_bdam8 [baidu/google]	1024	×	74.32	-	-	-	-	-

The results on DIOR-R are shown in the table below.(password: 0620)

Model ^(link)	Size ^(pixels)	TTA ^{(multi-scale/ rotate testing)}	OBB mAP^{test ^0.5 DIOR-R}	Speed ^{CPU b1 (ms)}	Speed ^{2080Ti b1 (ms)}	Speed ^{2080Ti b16 (ms)}	params ^(M)	FLOPs ^{@1024 (B)}
yolov5l_bdam5 [baidu/google]	1024	×	70.60	-	-	-	-	-

Table Notes (click to expand)

All checkpoints are trained to 300 epochs with COCO pre-trained checkpoints, default settings and hyperparameters.
mAP^test values are for single-model single-scale on DOTAv1.5 dataset.
Reproduce by python val.py --data 'data/dotav15_poly.yaml' --img 1024 --conf 0.01 --iou 0.4 --task 'test' --batch 16 --save-json
Speed averaged over DOTAv1.5 val_split_subsize1024_gap200 images using a 2080Ti gpu. NMS + pre-process times is included.
Reproduce by python val.py --data 'data/dotav15_poly.yaml' --img 1024 --task speed --batch 1

Please refer to install.md for installation and dataset preparation.

This repo is based on yolov5. Please see GetStart.md for the Oriented Detection basic usage.

I have used utility functions from other wonderful open-source projects. Espeicially thank the authors of:

  Name  : "杨刚"
  describe myself："菜鸟一枚"