This project is the PyTorch implementation of our accepted CVPR 2020 paper : forward and backward information retention for accurate binary neural networks. [PDF]
Bibtex:
@inproceedings{Qin:cvpr20,
author = {Haotong Qin and Ruihao Gong and Xianglong Liu and Mingzhu Shen and Ziran Wei and Fengwei Yu and Jingkuan Song},
title = {Forward and Backward Information Retention for Accurate Binary Neural Networks},
booktitle = {IEEE CVPR},
year = {2020},
}
IR-Net: We implement our IR-Net using Pytorch because of its high flexibility and powerful automatic differentiation mechanism. When constructing a binarized model, we simply replace the convolutional layers in the origin models with the binary convolutional layer binarized by our method.
Network Structures: We employ the widely-used network structures including VGG-Small, ResNet-20, ResNet-18 for CIFAR-10, and ResNet-18, ResNet-34 for ImageNet. To prove the versatility of our IR-Net, we evaluate it on both the normal structure and the Bi-Real structure of ResNet. All convolutional and fully-connected layers except the first and last one are binarized, and we select Hardtanh as our activation function instead of ReLU.
Initialization: Our IR-Net is trained from scratch (random initialization) without leveraging any pre-trained model. To evaluate our IR-Net on various network architectures, we mostly follow the hyper-parameter settings of their original papers. Among the experiments, we apply SGD as our optimization algorithm.
Dependencies
- Python 3.6
- Pytorch == 0.4.1
For the GPUs, we use a single NVIDIA GeForce 1080TI when training IR-Net on the CIFAR-10 dataset and 32 NVIDIA GeForce 1080TI when training IR-Net on the ImageNet dataset.
Accuracy:
CIFAR-10:
Topology | Bit-Width (W/A) | Accuracy (%) |
---|---|---|
ResNet-20 | 1 / 1 | 86.5 |
ResNet-20 | 1 / 32 | 90.8 |
VGG-Small | 1 / 1 | 90.4 |
ResNet-18 | 1 / 1 | 91.5 |
ImageNet:
Topology | Bit-Width (W/A) | Top-1 (%) | Top-5 (%) |
---|---|---|---|
ResNet-18 | 1 / 1 | 58.1 | 80.0 |
ResNet-18 | 1 / 32 | 66.5 | 86.8 |
ResNet-34 | 1 / 1 | 62.9 | 84.1 |
ResNet-34 | 1 / 32 | 70.4 | 89.5 |