A Study of lightweight Deep Learning library for resource-constrained embedded system

Deep learning libraries must be optimized to perform object detection with deep learning models in resource-constrained embedded hardware. To do this, we analyze hardware resource usage and energy consumption when executing the object detection. Base on the results of the analysis, we propose how to automatically find the optimal hardware control factors and adjust the model.

Associated Lab/Groups

Head of Project

Mailing address

  • Department of Computer Engineering
    Chungnam National University
    99, Daehak-ro, Yuseong-gu
    34134, Daejon, Korea



Hyungsin KimProfessor
Jaymin Lee (Alumni)Ph.D. Student
Sihyeong ParkPh.D. Student