Characterising Across-Stack Optimisations for Deep Convolutional Neural Networks

09/19/2018
by   Jack Turner, et al.
4

Convolutional Neural Networks (CNNs) are extremely computationally demanding, presenting a large barrier to their deployment on resource-constrained devices. Since such systems are where some of their most useful applications lie (e.g. obstacle detection for mobile robots, vision-based medical assistive technology), significant bodies of work from both machine learning and systems communities have attempted to provide optimisations that will make CNNs available to edge devices. In this paper we unify the two viewpoints in a Deep Learning Inference Stack and take an across-stack approach by implementing and evaluating the most common neural network compression techniques (weight pruning, channel pruning, and quantisation) and optimising their parallel execution with a range of programming approaches (OpenMP, OpenCL) and hardware architectures (CPU, GPU). We provide comprehensive Pareto curves to instruct trade-offs under constraints of accuracy, execution time, and memory space.

READ FULL TEXT

page 1

page 4

page 8

page 9

research
01/02/2017

Two-Bit Networks for Deep Learning on Resource-Constrained Embedded Devices

With the rapid proliferation of Internet of Things and intelligent edge ...
research
10/21/2020

Performance Prediction for Convolutional Neural Networks in Edge Devices

Running Convolutional Neural Network (CNN) based applications on edge de...
research
05/21/2020

TASO: Time and Space Optimization for Memory-Constrained DNN Inference

Convolutional neural networks (CNNs) are used in many embedded applicati...
research
07/27/2020

ALF: Autoencoder-based Low-rank Filter-sharing for Efficient Convolutional Neural Networks

Closing the gap between the hardware requirements of state-of-the-art co...
research
11/20/2016

LCNN: Lookup-based Convolutional Neural Network

Porting state of the art deep learning algorithms to resource constraine...
research
12/04/2014

Convolutional Neural Networks at Constrained Time Cost

Though recent advanced convolutional neural networks (CNNs) have been im...
research
05/10/2018

Towards Budget-Driven Hardware Optimization for Deep Convolutional Neural Networks using Stochastic Computing

Recently, Deep Convolutional Neural Network (DCNN) has achieved tremendo...

Please sign up or login with your details

Forgot password? Click here to reset