Video Analytics on IoT devices

TL;DR

This paper compares deep learning and traditional computer vision methods for video analytics on IoT devices, highlighting the most suitable approach based on performance and efficiency considerations.

Contribution

It provides a comparative analysis of DL-based and CV-based video analytics approaches specifically tailored for IoT devices.

Findings

DL approaches enable offline execution of large networks on IoT devices

Comparison shows strengths and weaknesses of DL and CV methods for IoT video analytics

Discussion identifies the most suitable approach for different IoT scenarios

Abstract

Deep Learning (DL) combined with advanced model optimization methods such as RC-NN and Edge2Train has enabled offline execution of large networks on the IoT devices. In this paper, we compare the modern Deep Learning (DL) based video analytics approaches with the standard Computer Vision (CV) based approaches and finally, discuss the best-suited approach for video analytics on IoT devices.