Deep Model Compression and Architecture Optimization for Embedded Systems: A Survey

Over the past, deep neural networks have proved to be an essential element for developing intelligent solutions. They have achieved remarkable performances at a cost of deeper layers and millions of parameters. Therefore utilising these networks on limited resource platforms for smart cameras is a challenging task. In this context, models need to be (i) accelerated and (ii) memory efficient without significantly compromising on performance. Numerous works have been done to obtain smaller, faster and accurate models. This paper presents a survey of methods suitable for porting deep neural networks on resource-limited devices, especially for smart cameras. These methods can be roughly divided in two main sections. In the first part, we present compression techniques. These techniques are categorized into: knowledge distillation, pruning, quantization, hashing, reduction of numerical precision and binarization. In the second part, we focus on architecture optimization. We introduce the methods to enhance networks structures as well as neural architecture search techniques. In each of their parts, we describe different methods, and analyse them. Finally, we conclude this paper with a discussion on these methods.

Automated summary

This paper surveys methods suitable for porting deep neural networks on resource-limited devices, especially for smart cameras, which can be roughly divided into compression techniques and architecture optimization. Compression techniques include knowledge distillation, pruning, quantization, hashing, reduction of numerical precision and binarization, while architecture optimization focuses on enhancing network structures and neural architecture search techniques.