Test code for BONN

Prepare

• install the latest PyTorch (1.1.0)
• download this code (pretrained model included)
• get the ImageNet validation set ready (you may need the script valprep.sh to pre-process the val set)

Check whether the weights and activation is binary

• run python weight_summary.py to check whether the weights are binary
• check the code in the module BinConv2d to see whether the input for convolution is binary

Evaluation

Modifications in the script evaluate_imagenet.sh:

• modify the PATH to your ImageNet dataset
• modify the batchsize (default: 256) according to your hardware (at least one GPU is requried)

Run the script evaluate_imagenet.sh and the accuracy on validation set is around 59.40.

Please cite

@inproceedings{gu2019bonn,
  title={Bayesian Optimized 1-bit CNNs},
  author={Gu, Jiaxin and Zhao, Junhe and Jiang, Xiaolong and Zhang, Baochang and Liu Jianzhuang and Guo, Guodong and Ji, Rongrong},
  booktitle={Proceedings of the IEEE International Conference on Computer Vision},
  year={2019}
}