Home AVLCode GitHub
image test

DSConv: Efficient Convolution Operator

This is the landing page for paper DSConv: Efficient Convolution Operator.


We introduce a variation of the convolutional layer called DSConv (Distribution Shifting Convolution) that can be readily substituted into standard neural network architectures and achieve both lower memory usage and higher computational speed. DSConv breaks down the traditional convolution kernel into two components: Variable Quantized Kernel (VQK), and Distribution Shifts. Lower memory usage and higher speeds are achieved by storing only integer values in the VQK, whilst preserving the same output as the original convolution by applying both kernel and channel based distribution shifts. We test DSConv in ImageNet on ResNet50 and 34, as well as AlexNet and MobileNet. We achieve a reduction in memory usage of up to 14x in the convolutional kernels and speed up operations of up to 10x by substituting floating point operations to integer operations. Furthermore, unlike other quantization approaches, our work allows for a degree of retraining to new tasks and datasets.


The code for reproducing results in the paper can be obtained from the GitHub repository.



author = {Gennari, Marcelo and Fawcett, Roger and Prisacariu, Victor Adrian},
title = {DSConv: Efficient Convolution Operator},
howpublished = {\url{https://arxiv.org/abs/1901.01928}},
year = {2019},
month = {January}

Plain text:

Marcelo Gennari, Roger Fawcett, Victor A. Prisacariu, “DSConv: Efficient Convolution Operator”, in arXiv:1901.01928, 2019