Details
Presenter(s)
Display Name
Hyunsung Kim
- Affiliation
-
AffiliationPOSTECH
- Country
Abstract
The electronic performance and tracking system (EPTS) device is one of the wearable devices widely used to gather sports data of athletes in real-time. This work demonstrates an adaptive sampling strategy controlled by the advanced on-device deep convolutional neural network (DCNN) operation, extending the lifetime of EPTS device even realized in a small form factor.