RBCN: Rectified Binary Convolutional Networks for Enhancing the Performance of 1-bit DCNNs [article]

Chunlei Liu and Wenrui Ding and Xin Xia and Yuan Hu and Baochang Zhang and Jianzhuang Liu and Bohan Zhuang and Guodong Guo
2019 arXiv   pre-print
Binarized convolutional neural networks (BCNNs) are widely used to improve memory and computation efficiency of deep convolutional neural networks (DCNNs) for mobile and AI chips based applications. However, current BCNNs are not able to fully explore their corresponding full-precision models, causing a significant performance gap between them. In this paper, we propose rectified binary convolutional networks (RBCNs), towards optimized BCNNs, by combining full-precision kernels and feature maps
more » ... to rectify the binarization process in a unified framework. In particular, we use a GAN to train the 1-bit binary network with the guidance of its corresponding full-precision model, which significantly improves the performance of BCNNs. The rectified convolutional layers are generic and flexible, and can be easily incorporated into existing DCNNs such as WideResNets and ResNets. Extensive experiments demonstrate the superior performance of the proposed RBCNs over state-of-the-art BCNNs. In particular, our method shows strong generalization on the object tracking task.
arXiv:1908.07748v2 fatcat:g2qkaqoeibhu7pw6q46sbs7eje