Next Article in Journal
Assessment of Vineyard Canopy Characteristics from Vigour Maps Obtained Using UAV and Satellite Imagery
Next Article in Special Issue
Study on Human Activity Recognition Using Semi-Supervised Active Transfer Learning
Previous Article in Journal
Synthesis of ZnAl2O4 and Evaluation of the Response in Propane Atmospheres of Pellets and Thick Films Manufactured with Powders of the Oxide
Previous Article in Special Issue
CCA: Cost-Capacity-Aware Caching for In-Memory Data Analytics Frameworks
Open AccessArticle

Accelerating On-Device Learning with Layer-Wise Processor Selection Method on Unified Memory

Human Enhancement & Assistive Technology Research Section, Artificial Intelligence Research Laboratory, Electronics Telecommunications Research Institute (ETRI), Daejeon 34129, Korea
*
Author to whom correspondence should be addressed.
Academic Editor: Sang-Woong Lee
Sensors 2021, 21(7), 2364; https://doi.org/10.3390/s21072364
Received: 28 February 2021 / Revised: 23 March 2021 / Accepted: 26 March 2021 / Published: 29 March 2021
Recent studies have applied the superior performance of deep learning to mobile devices, and these studies have enabled the running of the deep learning model on a mobile device with limited computing power. However, there is performance degradation of the deep learning model when it is deployed in mobile devices, due to the different sensors of each device. To solve this issue, it is necessary to train a network model specific to each mobile device. Therefore, herein, we propose an acceleration method for on-device learning to mitigate the device heterogeneity. The proposed method efficiently utilizes unified memory for reducing the latency of data transfer during network model training. In addition, we propose the layer-wise processor selection method to consider the latency generated by the difference in the processor performing the forward propagation step and the backpropagation step in the same layer. The experiments were performed on an ODROID-XU4 with the ResNet-18 model, and the experimental results indicate that the proposed method reduces the latency by at most 28.4% compared to the central processing unit (CPU) and at most 21.8% compared to the graphics processing unit (GPU). Through experiments using various batch sizes to measure the average power consumption, we confirmed that device heterogeneity is alleviated by performing on-device learning using the proposed method. View Full-Text
Keywords: deep learning acceleration; processor selection algorithm; on-device learning; acoustic scene classification; mobile devices deep learning acceleration; processor selection algorithm; on-device learning; acoustic scene classification; mobile devices
Show Figures

Figure 1

MDPI and ACS Style

Ha, D.; Kim, M.; Moon, K.; Jeong, C.Y. Accelerating On-Device Learning with Layer-Wise Processor Selection Method on Unified Memory. Sensors 2021, 21, 2364. https://doi.org/10.3390/s21072364

AMA Style

Ha D, Kim M, Moon K, Jeong CY. Accelerating On-Device Learning with Layer-Wise Processor Selection Method on Unified Memory. Sensors. 2021; 21(7):2364. https://doi.org/10.3390/s21072364

Chicago/Turabian Style

Ha, Donghee; Kim, Mooseop; Moon, KyeongDeok; Jeong, Chi Y. 2021. "Accelerating On-Device Learning with Layer-Wise Processor Selection Method on Unified Memory" Sensors 21, no. 7: 2364. https://doi.org/10.3390/s21072364

Find Other Styles
Note that from the first issue of 2016, MDPI journals use article numbers instead of page numbers. See further details here.

Article Access Map by Country/Region

1
Search more from Scilit
 
Search
Back to TopTop