Stroke Recovery of Handwritten Chinese Character using Fully Convolutional Networks

• Item type: SIG Technical Reports(1)
• 公開日: 2018-09-13
• タイトル: Stroke Recovery of Handwritten Chinese Character using Fully Convolutional Networks
• タイトル / 言語: en
• タイトル: Stroke Recovery of Handwritten Chinese Character using Fully Convolutional Networks
• 言語: eng
• キーワード / 主題Scheme: Other
• キーワード / 主題: ディスカッションセッション5
• 資源タイプ / 資源タイプ識別子: http://purl.org/coar/resource_type/c_18gh
• 資源タイプ: technical report
• 著者所属: Graduate School of Informatics, Kyoto University,
• 著者所属: Academic Center for Computing and Media Studies, Kyoto University,
• 著者所属: OMRON SINICX,Academic Center for Computing and Media Studies, Kyoto University,
• 著者所属(英): en
• 著者所属(英): Graduate School of Informatics, Kyoto University,
• 著者所属(英): en
• 著者所属(英): Academic Center for Computing and Media Studies, Kyoto University,
• 著者所属(英): en
• 著者所属(英): OMRON SINICX,Academic Center for Computing and Media Studies, Kyoto University,
• 著者名: Yujung, Wang; Motoharu, Sonogashira; Atsushi, Hashimoto; Masaaki, Iiyama
• 著者名(英): Yujung, Wang; Motoharu, Sonogashira; Atsushi, Hashimoto; Masaaki, Iiyama
• 論文抄録 / 内容記述タイプ: Other
• 論文抄録 / 内容記述: In this paper, we propose a method to recover strokes from offline handwritten Chinese characters. The proposed method employs multitask fully convolutional networks (FCN) to estimate writing order and direction of strokes from offline character images. Online data in CASIA database is used for training. Since the FCN produces discontinuous strokes, our method refines the estimation of the writing order by using a graph cut (GC). Experimental results with test dataset of CASIA-OLHWDB1.0 show the effectiveness of our method.
• 論文抄録(英) / 内容記述タイプ: Other
• 論文抄録(英) / 内容記述: In this paper, we propose a method to recover strokes from offline handwritten Chinese characters. The proposed method employs multitask fully convolutional networks (FCN) to estimate writing order and direction of strokes from offline character images. Online data in CASIA database is used for training. Since the FCN produces discontinuous strokes, our method refines the estimation of the writing order by using a graph cut (GC). Experimental results with test dataset of CASIA-OLHWDB1.0 show the effectiveness of our method.
• 書誌レコードID / 収録物識別子タイプ: NCID
• 書誌レコードID / 収録物識別子: AA11131797
• 書誌情報: 研究報告コンピュータビジョンとイメージメディア（CVIM）; 巻 2018-CVIM-213, 号 28, p. 1-6, 発行日 2018-09-13
• ISSN / 収録物識別子タイプ: ISSN
• ISSN / 収録物識別子: 2188-8701
• Notice: SIG Technical Reports are nonrefereed and hence may later appear in any journals, conferences, symposia, etc.
• 出版者 / 言語: ja
• 出版者: 情報処理学会