2023 Low-Power Computer Vision Challenge (LPCVC) Summary

TL;DR

The 2023 LPCVC focused on developing low-power, high-accuracy image segmentation solutions for UAVs on edge devices, highlighting innovative methods that balance resource constraints with performance.

Contribution

This paper introduces the 2023 LPCVC, emphasizing new approaches for efficient UAV image segmentation on edge devices, with detailed competition setup and winning methods.

Findings

Winners achieved high accuracy with low resource usage.

The competition attracted 60 international teams.

Solutions improved segmentation speed and accuracy.

Abstract

This article describes the 2023 IEEE Low-Power Computer Vision Challenge (LPCVC). Since 2015, LPCVC has been an international competition devoted to tackling the challenge of computer vision (CV) on edge devices. Most CV researchers focus on improving accuracy, at the expense of ever-growing sizes of machine models. LPCVC balances accuracy with resource requirements. Winners must achieve high accuracy with short execution time when their CV solutions run on an embedded device, such as Raspberry PI or Nvidia Jetson Nano. The vision problem for 2023 LPCVC is segmentation of images acquired by Unmanned Aerial Vehicles (UAVs, also called drones) after disasters. The 2023 LPCVC attracted 60 international teams that submitted 676 solutions during the submission window of one month. This article explains the setup of the competition and highlights the winners' methods that improve accuracy and shorten execution time.