Official Pytorch implementation of "PsyNet: Self-supervised Approach to Object Localization Using Point Symmetric Transformation"

This implementation is based on these repos.

• Pytorch Official ImageNet Example

Pre-trained checkpoints are now available

• Please check the "test" section.

• Minhyun Lee github (Equal contribution)
• Hyunjung Shim CV (Corresponding)

This paper is accepted by AAAI 2020. The pdf is available at https://aaai.org/Papers/AAAI/2020GB/AAAI-BaekK.206.pdf .

Existing co-localization techniques significantly lose performance over weakly or fully supervised methods in accuracy and inference time. In this paper, we overcome common drawbacks of co-localization techniques by utilizing self-supervised learning approach. The major technical contributions of the proposed method are two-fold. 1) We devise a new geometric transformation, namely point symmetric transformation and utilize its parameters as an artificial label for self-supervised learning. This new transformation can also play the role of region-drop based regularization. 2) We suggest a heat map extraction method for computing the heat map from the network trained by self-supervision, namely class-agnostic activation mapping. It is done by computing the spatial attention map. Based on extensive evaluations, we observe that the proposed method records new state-of-the-art performance in three fine-grained datasets for unsupervised object localization. Moreover, we show that the idea of the proposed method can be adopted in a modified manner to solve the weakly supervised object localization task. As a result, we outperform the current state-of-the-art technique in weakly supervised object localization by a significant gap.

• Code Refactoring

• python 3.6
• pytorch 1.0.0 or 1.1.0
• torchvision 0.2.2 or 0.3.0
• tqdm
• scipy
• PIL
• opencv-python (cv2)

• Download "CUB-200-2011" dataset from http://www.vision.caltech.edu/visipedia/CUB-200-2011.html
  • "Stanford Cars" from https://ai.stanford.edu/~jkrause/cars/car_dataset.html
  • "FGVC-aircraft" from http://www.robots.ox.ac.uk/~vgg/data/fgvc-aircraft/
• Extract in {"CUB", "stanfordCar", "AIRCRAFT/data"}.
• Then, donwload txt files at Link
• Place txt file under the dataset directory.
  • For CUB, ../data//CUB/CUB_200_2011/sizes.txt
  • For CARs, ../data/stanfordCar/devkit/cars_train_sizes.txt and ../data/stanfordCar/devkit/cars_test_sizes.txt
  • For Airplane, ../data/AIRCRAFT/data/images_train_size.txt and ../data/AIRCRAFT/data/images_val_size.txt and ../data/AIRCRAFT/data/images_test_size.txt
• Example directory hierarchy(CUB) :../data/CUB/CUB_200_2011

Project
|--- data
|    |--- CUB
|          |--- CUB_200_2011
|                    |--- images
|                    |      |--- 200 directories (001.Black ... ~ 200.Common ...)
|                    |--- sizes.txt
|--- PsyNet
     |--- models
     |        |--- networks.py
     |        |--- resnet.py
     |        |--- ...
     |--- main.py 
     |--- train.py
     | ...
     

python -W ignore main.py --dataset CUB --network vggcam16bn --tftypes OR

• Checkpoints

  • Link

• Performance

  • CUB - 83.78
  • CARs - 95.59
  • Aircraft - 96.61

• How to load .ckpt (Pre-trained checkpoint)

python -W ignore main.py --gpu 5 --dataset CUB --network vggcam16bn --tftypes OR --validation --load_model CUB_VGG16BN
python -W ignore main.py --gpu 5 --dataset AIRCRAFT --network vggcam16bn --tftypes OR --validation --load_model Aircraft_VGG16BN