Presentation | 2019-03-14 Aspect-Ratio-Preserving Multi-Patch Image Aesthetic Score Prediction. Lijie Wang, Xueting Wang, Toshihiko Yamasaki, Kiyoharu Aizawa, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | According to the spread of Social Networking Services (SNS), there is the increasing demand for automatically selecting, editing or generating aesthetic images and it raises importance of evaluating image aesthetics. We propose a multi-patch image aesthetic score prediction method with preserving original aspect-ratios of original images. In the experiment with AVA dataset containing 250,000 images, our approach outperforms the existing method in image aesthetic score prediction, especially improved the linear correlation coefficient between predicted and ground truth aesthetic scores by 0.05. Noticeably, significant decrease in prediction errors is observed with images having minor aspect-ratios. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | image aesthetics / score distribution / EMD / multi-patch learning / deep learning / CNN |
Paper # | IMQ2018-37,IE2018-121,MVE2018-68 |
Date of Issue | 2019-03-07 (IMQ, IE, MVE) |
Conference Information | |
Committee | IMQ / IE / MVE / CQ |
---|---|
Conference Date | 2019/3/14(2days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | Kagoshima University |
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 | Kenji Sugiyama(Seikei Univ.) / Takayuki Hamamoto(Tokyo Univ. of Science) / Kenji Mase(Nagoya Univ.) / Takanori Hayashi(Hiroshima Inst. of Tech.) |
Vice Chair | Toshiya Nakaguchi(Chiba Univ.) / Mitsuru Maeda(Canon) / Hideaki Kimata(NTT) / Kazuya Kodama(NII) / Masayuki Ihara(NTT) / Hideyuki Shimonishi(NEC) / Jun Okamoto(NTT) |
Secretary | Toshiya Nakaguchi(Nagoya Univ.) / Mitsuru Maeda(Sony) / Hideaki Kimata(KDDI Research) / Kazuya Kodama(Nagoya Univ.) / Masayuki Ihara(NTT) / Hideyuki Shimonishi(Kyushu Univ.) / Jun Okamoto(Nagoya Univ.) |
Assistant | Masaru Tsuchida(NTT) / Gosuke Ohashi(Shizuoka Univ.) / Kazuya Hayase(NTT) / Yasutaka Matsuo(NHK) / Satoshi Nishiguchi(Oosaka Inst. of Tech.) / Masanori Yokoyama(*) / Chikara Sasaki(KDDI Research) / Yoshiaki Nishikawa(NEC) / Ryo Yamamoto(UEC) |
Paper Information | |
Registration To | Technical Committee on Image Media Quality / Technical Committee on Image Engineering / Technical Committee on Media Experience and Virtual Environment / Technical Committee on Communication Quality |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | Aspect-Ratio-Preserving Multi-Patch Image Aesthetic Score Prediction. |
Sub Title (in English) | |
Keyword(1) | image aesthetics |
Keyword(2) | score distribution |
Keyword(3) | EMD |
Keyword(4) | multi-patch learning |
Keyword(5) | deep learning |
Keyword(6) | CNN |
1st Author's Name | Lijie Wang |
1st Author's Affiliation | The University of Tokyo(UTokyo) |
2nd Author's Name | Xueting Wang |
2nd Author's Affiliation | The University of Tokyo(UTokyo) |
3rd Author's Name | Toshihiko Yamasaki |
3rd Author's Affiliation | The University of Tokyo(UTokyo) |
4th Author's Name | Kiyoharu Aizawa |
4th Author's Affiliation | The University of Tokyo(UTokyo) |
Date | 2019-03-14 |
Paper # | IMQ2018-37,IE2018-121,MVE2018-68 |
Volume (vol) | vol.118 |
Number (no) | IMQ-500,IE-501,MVE-502 |
Page | pp.pp.85-90(IMQ), pp.85-90(IE), pp.85-90(MVE), |
#Pages | 6 |
Date of Issue | 2019-03-07 (IMQ, IE, MVE) |