English

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

Machine Learning 2021-03-08 v4 Computer Vision and Pattern Recognition

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.

Keywords

Cite

@article{arxiv.2010.00704,
  title  = {BCNN: A Binary CNN with All Matrix Ops Quantized to 1 Bit Precision},
  author = {Arthur J. Redfern and Lijun Zhu and Molly K. Newquist},
  journal= {arXiv preprint arXiv:2010.00704},
  year   = {2021}
}
R2 v1 2026-06-23T18:57:06.926Z