Presentation 2022-02-21
Domain Incremental Leaning with Adaptive Loss Functions
Takumi Kawashima, Go Irie, Daiki Ikami, Kiyoharu Aizawa,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) During domain incremental learning of image classification task, the distribution of images continually change, and models adapt themselves to images with new features while retaining past knowledge. There are very few methods specialized in it compared to class incremental learning, where new classes are incrementally added. In this work, we propose a simple method for domain incremental learning problems. Our proposed method is based on rehearsal with exemplars, and adaptively decide weights of loss functions. In our experiments, we use three datasets, one of which we made for this task.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) image classification / deep learning / catastrophic forgetting / domain incremental learning / distillation loss
Paper # ITS2021-30,IE2021-39
Date of Issue 2022-02-14 (ITS, IE)

Conference Information
Committee IE / ITS / ITE-AIT / ITE-ME / ITE-MMS
Conference Date 2022/2/21(2days)
Place (in Japanese) (See Japanese page)
Place (in English) Online
Topics (in Japanese) (See Japanese page)
Topics (in English) Image Processing, etc.
Chair Kazuya Kodama(NII) / Masahiro Fujii(Utsunomiya Univ.) / Hisaki Nate(Tokyo Polytechnic Univ.) / Hiroyuki Arai(Nippon Inst. of Tech.) / Kenji Machida(NHK)
Vice Chair Hiroyuki Bandoh(NTT) / Toshihiko Yamazaki(Univ. of Tokyo) / Kohei Ohno(Meiji Univ.) / Naohisa Hashimoto(AIST) / / Shogo Muramatsu(Niigata Univ.)
Secretary Hiroyuki Bandoh(KDDI Research) / Toshihiko Yamazaki(Nagoya Inst. of Tech.) / Kohei Ohno(Akita Prefectural Univ.) / Naohisa Hashimoto(NIT, Tsuruoka College) / / Shogo Muramatsu(NHK) / (Hokkaido Univ.)
Assistant Shunsuke Iwamura(NHK) / Shinobu Kudo(NTT) / Msataka Imao(Mitsubishi Electric) / Kenshi Saho(Toyama Prefectural Univ.) / Keiji Jimi(Gunma Univ.)

Paper Information
Registration To Technical Committee on Image Engineering / Technical Committee on Intelligent Transport Systems Technology / Technical Group on Artistic Image Technology / Technical Group on Media Engineering / Technical Group on Multi-media Storage
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Domain Incremental Leaning with Adaptive Loss Functions
Sub Title (in English)
Keyword(1) image classification
Keyword(2) deep learning
Keyword(3) catastrophic forgetting
Keyword(4) domain incremental learning
Keyword(5) distillation loss
1st Author's Name Takumi Kawashima
1st Author's Affiliation The University of Tokyo(UTokyo)
2nd Author's Name Go Irie
2nd Author's Affiliation Nippon Telegraph and Telephone Corporation(NTT)
3rd Author's Name Daiki Ikami
3rd Author's Affiliation Nippon Telegraph and Telephone Corporation(NTT)
4th Author's Name Kiyoharu Aizawa
4th Author's Affiliation The University of Tokyo(UTokyo)
Date 2022-02-21
Paper # ITS2021-30,IE2021-39
Volume (vol) vol.121
Number (no) ITS-373,IE-374
Page pp.pp.31-36(ITS), pp.31-36(IE),
#Pages 6
Date of Issue 2022-02-14 (ITS, IE)