Towards Approximate Edge Computing for Machine Learning & Digital Signal, Image and Video Processing (ApxEdgeC)

Type: International Partnership Project

Duration: from 2021 Sep 01 to 2023 Dec 31

Financed by: FCT-CAPES

Prime Contractor: R - INESC-ID Lisboa (Other) - Lisboa, Portugal

This cooperation project between the INESC-ID and groups of the Universidade Federal do Rio Grande do Sul (UFRGS) and the Universidade Católica de Pelotas(UCPel) aims to investigate approximate computing techniques in dedicated hardware to accelerate edge computing devices in the execution of machine learning algorithms and signal processing, images and videos, with greater energy efficiency. The scientific and technological goals targeted contributions in the development of new techniques and tools for systematic design of approximate hardware capable of crossing several abstraction layers up-to the application, aiming a rapid adoption of these techniques by the industry.

Partnerships

  • Universidade Católica de Pelotas (University) - Pelotas, Brazil
  • Universidade Federal do Rio Grande do Sul (University) - Porto Alegre, RS, Brazil

Principal Investigators

Members