BORex

How to run

1. Clone this repository.
2. Setup Python environment according to Setup section.
3. Download training/validation data from Pascal VOC dataset Development Kit.
   • http://host.robots.ox.ac.uk/pascal/VOC/ > The VOC2012 Challenge > Development Kit
4. Unzip .tar file, and place some directories to RISEBO/src/ as below.
   • Annotations as voc_annotation
   • JPEGImages as voc_image
   • SegmentationClass as voc_segmentation
5. Call python exec_borex.py or Run exec_borex.ipynb.

Setup

1. Upgrade pip.

   python -m pip3 install --upgrade pip
   

2. Install pytorch.

   Search .whl link according to environment at https://download.pytorch.org/whl/torch_stable.html.

   python -m pip install [.whl link]
   

   For example, https://download.pytorch.org/whl/cu116/torch-1.13.1%2Bcu116-cp310-cp310-win_amd64.whl for Win10 64bit, CUDA 11.6, Python3.10.

3. Install other packages.

   python -m pip install -r requirements.txt
   

   • requirements.txt

     torchvision
     scikit-learn
     matplotlib
     pyyaml
     tqdm
     scikit-image
     opencv-python
     

Confirmed environment

Python 3.9

• Python 3.9.13

• CUDA 11.6

  • NVIDIA Driver 528.02

• pip

  certifi==2022.12.7
  charset-normalizer==3.0.1
  colorama==0.4.6
  contourpy==1.0.7
  cycler==0.11.0
  fonttools==4.38.0
  idna==3.4
  imageio==2.25.0
  joblib==1.2.0
  kiwisolver==1.4.4
  matplotlib==3.6.3
  networkx==3.0
  numpy==1.24.1
  opencv-python==4.7.0.68
  packaging==23.0
  Pillow==9.4.0
  pyparsing==3.0.9
  python-dateutil==2.8.2
  PyWavelets==1.4.1
  PyYAML==6.0
  requests==2.28.2
  scikit-image==0.19.3
  scikit-learn==1.2.1
  scipy==1.10.0
  six==1.16.0
  threadpoolctl==3.1.0
  tifffile==2023.1.23.1
  torch @ https://download.pytorch.org/whl/cu116/torch-1.13.1%2Bcu116-cp39-cp39-win_amd64.whl
  torchvision==0.14.1
  tqdm==4.64.1
  typing_extensions==4.4.0
  urllib3==1.26.14
  

Python 3.10

• Python 3.10.8

• CUDA 11.6

  • NVIDIA Driver 528.02

• pip

  certifi==2022.12.7
  charset-normalizer==3.0.1
  colorama==0.4.6
  contourpy==1.0.7
  cycler==0.11.0
  fonttools==4.38.0
  idna==3.4
  imageio==2.25.0
  joblib==1.2.0
  kiwisolver==1.4.4
  matplotlib==3.6.3
  networkx==3.0
  numpy==1.24.1
  opencv-python==4.7.0.68
  packaging==23.0
  Pillow==9.4.0
  pyparsing==3.0.9
  python-dateutil==2.8.2
  PyWavelets==1.4.1
  PyYAML==6.0
  requests==2.28.2
  scikit-image==0.19.3
  scikit-learn==1.2.1
  scipy==1.10.0
  six==1.16.0
  threadpoolctl==3.1.0
  tifffile==2023.1.23.1
  torch @ https://download.pytorch.org/whl/cu116/torch-1.13.1%2Bcu116-cp310-cp310-win_amd64.whl
  torchvision==0.14.1
  tqdm==4.64.1
  typing_extensions==4.4.0
  urllib3==1.26.14
  

Reference

Atsushi Kikuchi, Kotaro Uchida, Masaki Waga, Kohei Suenaga: BOREx: Bayesian-Optimization-Based Refinement of Saliency Map for Image- and Video-Classification Models. ACCV (7) 2022: 274-290

@inproceedings{borex,
  author       = {Atsushi Kikuchi and
                  Kotaro Uchida and
                  Masaki Waga and
                  Kohei Suenaga},
  title        = {BOREx: Bayesian-Optimization-Based Refinement of Saliency Map for
                  Image- and Video-Classification Models},
  booktitle    = {{ACCV} 2022},
  pages        = {274--290},
  year         = {2022},
  url          = {https://doi.org/10.1007/978-3-031-26293-7\_17},
  doi          = {10.1007/978-3-031-26293-7\_17},
}