Presentation | 2022-03-11 A study on multi-task learning toward attractiveness estimation for food photography Mitsuaki Miyazaki, Keisuke Doman, Ichiro Ide, Yoshito Mekada, |
---|---|
PDF Download Page | ![]() |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | We report the results of a study on multi-task learning for the accuracy improvement of attractiveness estimation of food photography. We have been studying a way of improving the estimation accuracy by effectively using a small-scale image dataset. This paper proposes another approach based on multi-task learning for attractiveness estimation. Multi-task learning is a method that simultaneously learns multiple tasks within one model. Solving a task of interest together with its related tasks can improve the generalization performance of a trained model compared to solving each task independently. Thus, we expect that such a multi-task approach is effective for the attractiveness estimation of food photography. This report proposes a multi-task learning method that simultaneously solves multiple problems including attractiveness estimation as the main task and related estimation/classification problems as subtasks, and also confirm its effectiveness through evaluation experiments. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Food photography / attractiveness estimation / multi-task learning / regression analysis |
Paper # | IMQ2021-68,IE2021-130,MVE2021-97 |
Date of Issue | 2022-03-02 (IMQ, IE, MVE) |
Conference Information | |
Committee | CQ / IMQ / MVE / IE |
---|---|
Conference Date | 2022/3/9(3days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | Online (Zoom) |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | Media of five senses, Multimedia, Media experience, Picture codinge, Image media quality, Network,quality and reliability, etc |
Chair | Jun Okamoto(NTT) / Kenya Uomori(Osaka Univ.) / Masayuki Ihara(RIKEN) / Kazuya Kodama(NII) |
Vice Chair | Takefumi Hiraguri(Nippon Inst. of Tech.) / Gou Hasegawa(Tohoku Univ.) / Mitsuru Maeda(Canon) / Kiyoshi Kiyokawa(NAIST) / Hiroyuki Bandoh(NTT) / Toshihiko Yamazaki(Univ. of Tokyo) |
Secretary | Takefumi Hiraguri(NTT) / Gou Hasegawa(Ritsumeikan Univ.) / Mitsuru Maeda(Nagoya Univ.) / Kiyoshi Kiyokawa(NTT) / Hiroyuki Bandoh(Oosaka Inst. of Tech.) / Toshihiko Yamazaki(NTT) |
Assistant | Yoshiaki Nishikawa(NEC) / Ryoichi Kataoka(KDDI Research) / Kimiko Kawashima(NTT) / Masato Tsukada(NEC) / Takashi Yamazoe(Seikei Univ.) / Naoya Isoyama(NAIST) / Takenori Hara(DNP) / Mitsuhiro Goto(NTT) / Shunsuke Iwamura(NHK) / Shinobu Kudo(NTT) |
Paper Information | |
Registration To | Technical Committee on Communication Quality / Technical Committee on Image Media Quality / Technical Committee on Media Experience and Virtual Environment / Technical Committee on Image Engineering |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | A study on multi-task learning toward attractiveness estimation for food photography |
Sub Title (in English) | |
Keyword(1) | Food photography |
Keyword(2) | attractiveness estimation |
Keyword(3) | multi-task learning |
Keyword(4) | regression analysis |
1st Author's Name | Mitsuaki Miyazaki |
1st Author's Affiliation | Chukyo University(Chukyo Univ.) |
2nd Author's Name | Keisuke Doman |
2nd Author's Affiliation | Chukyo University(Chukyo Univ.) |
3rd Author's Name | Ichiro Ide |
3rd Author's Affiliation | Nagoya University(Nagoya Univ.) |
4th Author's Name | Yoshito Mekada |
4th Author's Affiliation | Chukyo University(Chukyo Univ.) |
Date | 2022-03-11 |
Paper # | IMQ2021-68,IE2021-130,MVE2021-97 |
Volume (vol) | vol.121 |
Number (no) | IMQ-420,IE-422,MVE-423 |
Page | pp.pp.306-311(IMQ), pp.306-311(IE), pp.306-311(MVE), |
#Pages | 6 |
Date of Issue | 2022-03-02 (IMQ, IE, MVE) |