Presentation 2019-06-17
Embedded AI implementation
Tadaaki Shiraishi, Koji Kinoshita,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In order to realize CNN on edge terminals and realize inference processing such as image recognition, it is important to reduce power consumption and realize price reduction while maintaining processing speed and recognition accuracy. In this paper, we realized the weight reduction of CNN, implemented it in FPGA, and compared it with CPU and embedded system GPU, and report the result. When small-scale CNNs that recognize handwritten numbers were implemented on each device and compared, FPGA implementation was good in terms of price, power consumption, and processing speed.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) FPGA, / CNN / Weight Reduction / Edge Computing
Paper # NC2019-10
Date of Issue 2019-06-10 (NC)

Conference Information
Committee NC / IBISML / IPSJ-MPS / IPSJ-BIO
Conference Date 2019/6/17(3days)
Place (in Japanese) (See Japanese page)
Place (in English) Okinawa Institute of Science and Technology
Topics (in Japanese) (See Japanese page)
Topics (in English) Neurocomputing, Machine Learning Approach to Biodata Mining, and General
Chair Hayaru Shouno(UEC) / Hisashi Kashima(Kyoto Univ.) / Masakazu Sekijima(Tokyo Tech) / Hiroyuki Kurata(Kyutech)
Vice Chair Kazuyuki Samejima(Tamagawa Univ) / Masashi Sugiyama(Univ. of Tokyo) / Koji Tsuda(Univ. of Tokyo)
Secretary Kazuyuki Samejima(NAIST) / Masashi Sugiyama(NTT) / Koji Tsuda(Nagoya Inst. of Tech.) / (AIST) / (Nagoya Univ.)
Assistant Takashi Shinozaki(NICT) / Ken Takiyama(TUAT) / Tomoharu Iwata(NTT) / Shigeyuki Oba(Kyoto Univ.)

Paper Information
Registration To Technical Committee on Neurocomputing / Technical Committee on Infomation-Based Induction Sciences and Machine Learning / IPSJ Special Interest Group on Mathematical Modeling and Problem Solving / IPSJ Special Interest Group on Bioinformatics and Genomics
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Embedded AI implementation
Sub Title (in English)
Keyword(1) FPGA,
Keyword(2) CNN
Keyword(3) Weight Reduction
Keyword(4) Edge Computing
1st Author's Name Tadaaki Shiraishi
1st Author's Affiliation Mitsubishi Electric Micro-Computer Application Software Co.,Ltd(MMS)
2nd Author's Name Koji Kinoshita
2nd Author's Affiliation Ehime University(Ehime Univ)
Date 2019-06-17
Paper # NC2019-10
Volume (vol) vol.119
Number (no) NC-88
Page pp.pp.35-40(NC),
#Pages 6
Date of Issue 2019-06-10 (NC)