神戸大学附属図書館デジタルアーカイブ
入力補助
English
カテゴリ
学内刊行物
ランキング
アクセスランキング
ダウンロードランキング
https://hdl.handle.net/20.500.14094/90007755
このアイテムのアクセス数:
190
件
(
2024-07-28
00:17 集計
)
閲覧可能ファイル
ファイル
フォーマット
サイズ
閲覧回数
説明
90007755 (fulltext)
pdf
3.91 MB
29
メタデータ
ファイル出力
メタデータID
90007755
アクセス権
open access
出版タイプ
Version of Record
タイトル
Multi-Category Image Super-Resolution with Convolutional Neural Network and Multi-Task Learning
著者
URAZOE, Kazuya ; KUROKI, Nobutaka ; KATO, Yu ; OHTANI, Shinya ; HIROSE, Tetsuya ; NUMA, Masahiro
著者名
URAZOE, Kazuya
著者ID
A0918
研究者ID
1000090273763
KUID
https://kuid-rm-web.ofc.kobe-u.ac.jp/search/detail?systemId=edb7c57261d5dfdf520e17560c007669
著者名
KUROKI, Nobutaka
黒木, 修隆
クロキ, ノブタカ
所属機関名
工学研究科
著者名
KATO, Yu
著者名
OHTANI, Shinya
著者名
HIROSE, Tetsuya
著者ID
A0131
研究者ID
1000060188787
KUID
https://kuid-rm-web.ofc.kobe-u.ac.jp/search/detail?systemId=05d1d70f3dd4e277520e17560c007669
著者名
NUMA, Masahiro
沼, 昌宏
ヌマ, マサヒロ
所属機関名
工学研究科
言語
English (英語)
収録物名
IEICE Transactions on Information and Systems
巻(号)
E104.D(1)
ページ
183-193
出版者
Institute of Electronics, Information and Communication Engineers(IEICE)
刊行日
2021-01-01
公開日
2021-01-14
抄録
This paper presents an image super-resolution technique using a convolutional neural network (CNN) and multi-task learning for multiple image categories. The image categories include natural, manga, and text images. Their features differ from each other. However, several CNNs for super-resolution are trained with a single category. If the input image category is different from that of the training images, the performance of super-resolution is degraded. There are two possible solutions to manage multi-categories with conventional CNNs. The first involves the preparation of the CNNs for every category. This solution, however, requires a category classifier to select an appropriate CNN. The second is to learn all categories with a single CNN. In this solution, the CNN cannot optimize its internal behavior for each category. Therefore, this paper presents a super-resolution CNN architecture for multiple image categories. The proposed CNN has two parallel outputs for a high-resolution image and a category label. The main CNN for the high-resolution image is a normal three convolutional layer-architecture, and the sub neural network for the category label is branched out from its middle layer and consists of two fully-connected layers. This architecture can simultaneously learn the high-resolution image and its category using multi-task learning. The category information is used for optimizing the super-resolution. In an applied setting, the proposed CNN can automatically estimate the input image category and change the internal behavior. Experimental results of 2× image magnification have shown that the average peak signal-to-noise ratio for the proposed method is approximately 0.22 dB higher than that for the conventional super-resolution with no difference in processing time and parameters. We have ensured that the proposed method is useful when the input image category is varying.
キーワード
super-resolution
resolution enhancement
convolutional neural network
multi-task learning
deep learning
カテゴリ
工学研究科
学術雑誌論文
権利
© 2021 The Institute of Electronics, Information and Communication Engineers
関連情報
DOI
https://doi.org/10.1587/transinf.2020EDP7054
NAID
130007965063
CiNiiで表示
詳細を表示
資源タイプ
journal article
ISSN
0916-8532
OPACで所蔵を検索
CiNiiで学外所蔵を検索
eISSN
1745-1361
OPACで所蔵を検索
CiNiiで学外所蔵を検索
ホームへ戻る