Performance Evaluation of Pipeline-Based Processing for the Caffe Deep Learning Framework

Ayae ICHINOSE; Atsuko TAKEFUSA; Hidemoto NAKADA; Masato OGUCHI

doi:10.1587/transinf.2017DAP0015

Performance Evaluation of Pipeline-Based Processing for the Caffe Deep Learning Framework

Ayae ICHINOSE, Atsuko TAKEFUSA, Hidemoto NAKADA, Masato OGUCHI

Full Text Views

0

Cite this

Summary :

Many life-log analysis applications, which transfer data from cameras and sensors to a Cloud and analyze them in the Cloud, have been developed as the use of various sensors and Cloud computing technologies has spread. However, difficulties arise because of the limited network bandwidth between such sensors and the Cloud. In addition, sending raw sensor data to a Cloud may introduce privacy issues. Therefore, we propose a pipelined method for distributed deep learning processing between sensors and the Cloud to reduce the amount of data sent to the Cloud and protect the privacy of users. In this study, we measured the processing times and evaluated the performance of our method using two different datasets. In addition, we performed experiments using three types of machines with different performance characteristics on the client side and compared the processing times. The experimental results show that the accuracy of deep learning with coarse-grained data is comparable to that achieved with the default parameter settings, and the proposed distributed processing method has performance advantages in cases of insufficient network bandwidth between realistic sensors and a Cloud environment. In addition, it is confirmed that the process that most affects the overall processing time varies depending on the machine performance on the client side, and the most efficient distribution method similarly differs.

Publication: IEICE TRANSACTIONS on Information Vol.E101-D No.4 pp.1042-1052

Publication Date: 2018/04/01

Publicized: 2018/01/18

Online ISSN: 1745-1361

DOI: 10.1587/transinf.2017DAP0015

Type of Manuscript: Special Section PAPER (Special Section on Data Engineering and Information Management)

Category

Authors

Ayae ICHINOSE
  Ochanomizu University
Atsuko TAKEFUSA
  National Institute of Informatics
Hidemoto NAKADA
  National Institute of Advanced Industrial Science and Technology (AIST)
Masato OGUCHI
  Ochanomizu University

Keyword

deep learning, machine learning, distributed processing, cloud computing, life-log analysis

Cite this

Copy

Ayae ICHINOSE, Atsuko TAKEFUSA, Hidemoto NAKADA, Masato OGUCHI, "Performance Evaluation of Pipeline-Based Processing for the Caffe Deep Learning Framework" in IEICE TRANSACTIONS on Information, vol. E101-D, no. 4, pp. 1042-1052, April 2018, doi: 10.1587/transinf.2017DAP0015.
Abstract: Many life-log analysis applications, which transfer data from cameras and sensors to a Cloud and analyze them in the Cloud, have been developed as the use of various sensors and Cloud computing technologies has spread. However, difficulties arise because of the limited network bandwidth between such sensors and the Cloud. In addition, sending raw sensor data to a Cloud may introduce privacy issues. Therefore, we propose a pipelined method for distributed deep learning processing between sensors and the Cloud to reduce the amount of data sent to the Cloud and protect the privacy of users. In this study, we measured the processing times and evaluated the performance of our method using two different datasets. In addition, we performed experiments using three types of machines with different performance characteristics on the client side and compared the processing times. The experimental results show that the accuracy of deep learning with coarse-grained data is comparable to that achieved with the default parameter settings, and the proposed distributed processing method has performance advantages in cases of insufficient network bandwidth between realistic sensors and a Cloud environment. In addition, it is confirmed that the process that most affects the overall processing time varies depending on the machine performance on the client side, and the most efficient distribution method similarly differs.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2017DAP0015/_p

Copy

@ARTICLE{e101-d_4_1042,
author={Ayae ICHINOSE, Atsuko TAKEFUSA, Hidemoto NAKADA, Masato OGUCHI, },
journal={IEICE TRANSACTIONS on Information},
title={Performance Evaluation of Pipeline-Based Processing for the Caffe Deep Learning Framework},
year={2018},
volume={E101-D},
number={4},
pages={1042-1052},
abstract={Many life-log analysis applications, which transfer data from cameras and sensors to a Cloud and analyze them in the Cloud, have been developed as the use of various sensors and Cloud computing technologies has spread. However, difficulties arise because of the limited network bandwidth between such sensors and the Cloud. In addition, sending raw sensor data to a Cloud may introduce privacy issues. Therefore, we propose a pipelined method for distributed deep learning processing between sensors and the Cloud to reduce the amount of data sent to the Cloud and protect the privacy of users. In this study, we measured the processing times and evaluated the performance of our method using two different datasets. In addition, we performed experiments using three types of machines with different performance characteristics on the client side and compared the processing times. The experimental results show that the accuracy of deep learning with coarse-grained data is comparable to that achieved with the default parameter settings, and the proposed distributed processing method has performance advantages in cases of insufficient network bandwidth between realistic sensors and a Cloud environment. In addition, it is confirmed that the process that most affects the overall processing time varies depending on the machine performance on the client side, and the most efficient distribution method similarly differs.},
keywords={},
doi={10.1587/transinf.2017DAP0015},
ISSN={1745-1361},
month={April},}

Copy

TY - JOUR
TI - Performance Evaluation of Pipeline-Based Processing for the Caffe Deep Learning Framework
T2 - IEICE TRANSACTIONS on Information
SP - 1042
EP - 1052
AU - Ayae ICHINOSE
AU - Atsuko TAKEFUSA
AU - Hidemoto NAKADA
AU - Masato OGUCHI
PY - 2018
DO - 10.1587/transinf.2017DAP0015
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E101-D
IS - 4
JA - IEICE TRANSACTIONS on Information
Y1 - April 2018
AB - Many life-log analysis applications, which transfer data from cameras and sensors to a Cloud and analyze them in the Cloud, have been developed as the use of various sensors and Cloud computing technologies has spread. However, difficulties arise because of the limited network bandwidth between such sensors and the Cloud. In addition, sending raw sensor data to a Cloud may introduce privacy issues. Therefore, we propose a pipelined method for distributed deep learning processing between sensors and the Cloud to reduce the amount of data sent to the Cloud and protect the privacy of users. In this study, we measured the processing times and evaluated the performance of our method using two different datasets. In addition, we performed experiments using three types of machines with different performance characteristics on the client side and compared the processing times. The experimental results show that the accuracy of deep learning with coarse-grained data is comparable to that achieved with the default parameter settings, and the proposed distributed processing method has performance advantages in cases of insufficient network bandwidth between realistic sensors and a Cloud environment. In addition, it is confirmed that the process that most affects the overall processing time varies depending on the machine performance on the client side, and the most efficient distribution method similarly differs.
ER -