# How Does Batch Normalization Help Binary Training?

arXiv: https://arxiv.org/abs/1909.09139 · HTML: https://mlsanity.com/paper/1909.09139/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 18 Sep 2019.

## Built on by (most cited first, 8 of 8)

- [Enabling Binary Neural Network Training on the Edge](https://mlsanity.com/paper/2102.04270/) (2021)
- [Bit Error Tolerance Metrics for Binarized Neural Networks](https://mlsanity.com/paper/2102.01344/) (2021)
- [Universal Approximation Theorems of Fully Connected Binarized Neural Networks](https://mlsanity.com/paper/2102.02631/) (2021)
- [Lightweight Jet Reconstruction and Identification as an Object Detection Task](https://mlsanity.com/paper/2202.04499/) (2022)
- [PR-CIM: a Variation-Aware Binary-Neural-Network Framework for Process-Resilient Computation-in-memory](https://mlsanity.com/paper/2110.09962/) (2021)
- [On-Sensor Binarized Fully Convolutional Neural Network with A Pixel Processor Array](https://mlsanity.com/paper/2202.00836/) (2022)
- [HEP-BNN: A Framework for Finding Low-Latency Execution Configurations of BNNs on Heterogeneous Multiprocessor Platforms](https://mlsanity.com/paper/2301.05126/) (2023)
- [Bringing BNNs to Fast Event Processing](https://mlsanity.com/paper/2610.09873/) (2026)

## Abstract (arXiv)

Binary Neural Networks (BNNs) are difficult to train, and suffer from drop of accuracy. It appears in practice that BNNs fail to train in the absence of Batch Normalization (BatchNorm) layer. We find the main role of BatchNorm is to avoid exploding gradients in the case of BNNs. This finding suggests that the common initialization methods developed for full-precision networks are irrelevant to BNNs. We build a theoretical study on the role of BatchNorm in binary training, backed up by numerical experiments.