Awesome

ProbIOU Loss

Base code here

Losses

Losses can be chosen with the --losstype option in the arguments in train.py file. The valid options are currently: [Iou|Giou|Diou|Ciou|SmoothL1|Piou].

Fold-Structure

The fold structure as follow:

• config/
  • config.py
  • init.py
• data/
  • init.py
  • VOC.py
  • VOCdevkit/
• model/
  • build_ssd.py
  • init.py
  • backbone/
  • neck/
  • head/
  • utils/
• utils/
  • box/
  • detection/
  • loss/
  • init.py
• tools/
  • train.py
  • eval.py
  • test.py
• work_dir/

Environment

• conda create --name probiou-ssd python=3.6
• conda activate probiou-ssd
• pip install -r requirements.txt

Datasets

• PASCAL VOC:Download VOC2007, VOC2012 dataset, then put VOCdevkit in the data directory or run get_voc_dataset.sh in data folder

Training

Training VOC

python tools/train.py --loss <loss_type> --work_name <save_path>

Also yo can activate python -m visdom.server in an additional tmux window to track the losses.

Evaluation

• To evaluate a trained network:

python tools/ap.py --trained_model {your_weight_address} --ProbIoU [True/False]

For example: (the output is AP50, AP75 and AP of our CIoU loss)

Results:
0.033
0.015
0.009
0.011
0.008
0.083
0.044
0.042
0.004
0.014
0.026
0.034
0.010
0.006
0.009
0.006
0.009
0.013
0.106
0.011
0.025
~~~~~~~~

--------------------------------------------------------------
Results computed with the **unofficial** Python eval code.
Results should be very close to the official MATLAB eval code.
--------------------------------------------------------------
0.7884902583981603 0.5615516772893671 0.5143832356646468

Test

• To test a trained network:

python test.py -- trained_model {your_weight_address}

if you want to visual the box, you can add the command --visbox True(default False)

Cite our work

@article{Murrugarra_Llerena_2024,
   title={Probabilistic Intersection-Over-Union for Training and Evaluation of Oriented Object Detectors},
   volume={33},
   ISSN={1941-0042},
   url={http://dx.doi.org/10.1109/TIP.2023.3348697},
   DOI={10.1109/tip.2023.3348697},
   journal={IEEE Transactions on Image Processing},
   publisher={Institute of Electrical and Electronics Engineers (IEEE)},
   author={Murrugarra-Llerena, Jeffri and Kirsten, Lucas N. and Zeni, Luis Felipe and Jung, Claudio R.},
   year={2024},
   pages={671–681} }

FOR QUESTION

email me at: jeffri.mllerena@inf.ufrgs.br