Edinburgh Research Explorer

Algorithmic Performance-Accuracy Trade-off in 3D Vision Applications Using HyperMapper

Research output: Chapter in Book/Report/Conference proceedingConference contribution

  • Luigi Nardi
  • Bruno Bodin
  • Sajad Saeedi
  • Emanuele Vespa
  • Andrew J Davison
  • Paul H. J. Kelly

Related Edinburgh Organisations

Open Access permissions

Open

Documents

http://hdl.handle.net/10044/1/45399
Original languageEnglish
Title of host publicationThe Twelfth International Workshop on Automatic Performance Tuning 2017
Number of pages10
StatePublished - 2 Jun 2017
Event12th International Workshop on Automatic Performance Tuning held in conjunction with 31th IEEE International Parallel & Distributed Processing Symposium (iWAP2017) - Orlando, United States
Duration: 2 Jun 2017 → …
Conference number: 2017
http://www.iwapt.org/2017/

Workshop

Workshop12th International Workshop on Automatic Performance Tuning held in conjunction with 31th IEEE International Parallel & Distributed Processing Symposium (iWAP2017)
Abbreviated titleiWAPT
CountryUnited States
CityOrlando
Period2/06/17 → …
Internet address

Abstract

In this paper we investigate an emerging application, 3D scene understanding, likely to be significant in the mobile space in the near future. The goal of this exploration is to reduce execution time while meeting our quality of result objectives. In previous work we showed for the first time that it is possible to map this application to power constrained embedded systems, highlighting that decision choices made at the algorithmic design-level have the most impact. As the algorithmic design space is too large to be exhaustively evaluated, we use a previously introduced multi-objective Random Forest Active Learning prediction framework dubbed HyperMapper, to find good algorithmic designs. We show that HyperMapper generalizes on a recent cutting edge 3D scene understanding algorithm and on a modern GPU-based computer architecture. HyperMapper is able to beat an expert human hand-tuning the algorithmic parameters of the class of Computer Vision applications taken under consideration in this paper automatically. In addition, we use crowd-sourcing using a 3D scene understanding Android app to show that the Pareto front obtained on an embedded system can be used to accelerate the same application on all the 83 smart-phones and tablets crowd-sourced with speedups ranging from 2 to over 12.

ID: 61149501