BCNN: A Binary CNN with All Matrix Ops Quantized to 1 Bit Precision

2020/10/01 by Redfern, Arthur J., Zhu, Lijun, Newquist, Molly K.
#Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Machine Learning (cs.LG)

paper · doi:10.48550/arxiv.2010.00704

Abstract

This paper describes a CNN where all CNN style 2D convolution operations that lower to matrix matrix multiplication are fully binary. The network is derived from a common building block structure that is consistent with a constructive proof outline showing that binary neural networks are universal function approximators. 71.24% top 1 accuracy on the 2012 ImageNet validation set was achieved with a 2 step training procedure and implementation strategies optimized for binary operands are provided.

Related

• FTBNN: Rethinking Non-linearity for 1-bit CNNs and Going Beyond
• Back to Simplicity: How to Train Accurate BNNs from Scratch?
• Join the High Accuracy Club on ImageNet with A Binary Neural Network Ticket
• BCNN: Binary Complex Neural Network
• BD-Net: Has Depth-Wise Convolution Ever Been Applied in Binary Neural Networks?