Close

Session

Research Manuscript: Little Devices Doing Big Things: ML Techniques for Resource-Constrained Embedded Devices
Event TypeResearch Manuscript
Keywords
Embedded System Design Methodologies
Topics
Embedded Systems
TimeWednesday, July 13th10:30am - 12:00pm PDT
Location3004, Level 3
DescriptionResource-constrained computing remains an important challenge in embedded systems, especially as these systems become more complex and feature-rich. This session covers cutting-edge research to enable efficient computation in emerging embedded systems applications. The first paper proposes a lightweight hardware-aware differentiable network architecture search framework to find the best architecture in a single search pass. The second paper proposes a method to reduce the energy consumption of sensor fusion in difficult autonomous driving contexts. The third paper explores the use of human emotions to facilitate energy-efficient hardware system management in edge devices.