Efficient Winograd Convolution via Integer Arithmetic

01/07/2019
by   Lingchuan Meng, et al.
0

Convolution is the core operation for many deep neural networks. The Winograd convolution algorithms have been shown to accelerate the widely-used small convolution sizes. Quantized neural networks can effectively reduce model sizes and improve inference speed, which leads to a wide variety of kernels and hardware accelerators that work with integer data. The state-of-the-art Winograd algorithms pose challenges for efficient implementation and execution by the integer kernels and accelerators. We introduce a new class of Winograd algorithms by extending the construction to the field of complex and propose optimizations that reduce the number of general multiplications. The new algorithm achieves an arithmetic complexity reduction of 3.13x over the direct method and an efficiency gain up to 17.37% over the rational algorithms. Furthermore, we design and implement an integer-based filter scaling scheme to effectively reduce the filter bit width by 30.77% without any significant accuracy loss.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
07/23/2020

Efficient Residue Number System Based Winograd Convolution

Prior research has shown that Winograd algorithm can reduce the computat...
research
09/26/2022

Going Further With Winograd Convolutions: Tap-Wise Quantization for Efficient Inference on 4x4 Tile

Most of today's computer vision pipelines are built around deep neural n...
research
04/08/2023

Arithmetic Intensity Balancing Convolution for Hardware-aware Efficient Block Design

As deep learning advances, edge devices and lightweight neural networks ...
research
09/27/2018

Scalar Arithmetic Multiple Data: Customizable Precision for Deep Neural Networks

Quantization of weights and activations in Deep Neural Networks (DNNs) i...
research
11/29/2020

XpulpNN: Enabling Energy Efficient and Flexible Inference of Quantized Neural Network on RISC-V based IoT End Nodes

This work introduces lightweight extensions to the RISC-V ISA to boost t...
research
04/23/2020

Quantaized Winograd/Toom-Cook Convolution for DNNs: Beyond Canonical Polynomials Base

The problem how to speed up the convolution computations in Deep Neural ...
research
04/19/2023

Neural Network Quantisation for Faster Homomorphic Encryption

Homomorphic encryption (HE) enables calculating on encrypted data, which...

Please sign up or login with your details

Forgot password? Click here to reset