Presentation 2020-03-08
Proposal of Data Augmentation Methods for Echo Sounder Image Analysis
Min Jie, Soichiro Yokoyama, Tomohisa Yamashita, Hidenori Kawamura,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Data augmentation plays an important role in deep learning. Recently, RandAugment have been proposed as effective augmentation methods. In aquaculture, echo sounder image analysis is used for catching specific kind of fish to improve annual catches. Therefore, it remains significant to research for effects of different methods and find suitable augmentation method for echo sounder image analysis. This paper aims at finding effective settings for RandAugment by comparing effects of different transformation methods and figure out fitness of basic methods used in RandAugment for sonar image according to their variations with original images. The experiment results show improvement on recall rate and f1 score of distinguishing tuna in echo sounder images with suitable transformation methods.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Echo sounder imageData AugmentationRandAugmentDeep learning
Paper # AI2019-54
Date of Issue 2020-03-01 (AI)

Conference Information
Committee AI / IPSJ-ICS / JSAI-SAI / JSAI-DOCMAS / JSAI-KBS
Conference Date 2020/3/8(2days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English) Workshop of Social System and Information Technology (WSSIT20)
Chair Naoki Fukuta(Shizuoka Univ.)
Vice Chair Yuichi Sei(Univ. of Electro-Comm.) / Yuko Sakurai(AIST)
Secretary Yuichi Sei(Osaka Univ.) / Yuko Sakurai(Tokyo Univ. of Agriculture and Technology)
Assistant

Paper Information
Registration To Technical Committee on Artificial Intelligence and Knowledge-Based Processing / Special Interest Group on Intelligence and Complex Systems / Special Interest Group on Society and Artificial Intelligence / Special Interest Group on Data Oriented Constructive Mining and Simulation / Special Interest Group on Knowledge-Based Systems
Language ENG
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Proposal of Data Augmentation Methods for Echo Sounder Image Analysis
Sub Title (in English)
Keyword(1) Echo sounder imageData AugmentationRandAugmentDeep learning
1st Author's Name Min Jie
1st Author's Affiliation Hokkaido University(Hokudai)
2nd Author's Name Soichiro Yokoyama
2nd Author's Affiliation Hokkaido University(Hokudai)
3rd Author's Name Tomohisa Yamashita
3rd Author's Affiliation Hokkaido University(Hokudai)
4th Author's Name Hidenori Kawamura
4th Author's Affiliation Hokkaido University(Hokudai)
Date 2020-03-08
Paper # AI2019-54
Volume (vol) vol.119
Number (no) AI-469
Page pp.pp.1-5(AI),
#Pages 5
Date of Issue 2020-03-01 (AI)