An Overview of Datatype Quantization Techniques for Convolutional Neural Networks

Ali Athar

An Overview of Datatype Quantization Techniques for Convolutional Neural Networks

Neural and Evolutionary Computing 2018-08-24 v1

Authors: Ali Athar

Abstract

Convolutional Neural Networks (CNNs) are becoming increasingly popular due to their superior performance in the domain of computer vision, in applications such as objection detection and recognition. However, they demand complex, power-consuming hardware which makes them unsuitable for implementation on low-power mobile and embedded devices. In this paper, a description and comparison of various techniques is presented which aim to mitigate this problem. This is primarily achieved by quantizing the floating-point weights and activations to reduce the hardware requirements, and adapting the training and inference algorithms to maintain the network's performance.

Keywords

convolutional neural network deep neural network acceleration quantization

Cite

@article{arxiv.1808.07530,
  title  = {An Overview of Datatype Quantization Techniques for Convolutional Neural Networks},
  author = {Ali Athar},
  journal= {arXiv preprint arXiv:1808.07530},
  year   = {2018}
}

Comments

4 pages, 2 figures

Related papers

View all related →