Continuous Variables Estimation Through Classification Networks Ensembles

• Item type: SIG Technical Reports(1)
• 公開日: 2020-03-09
• タイトル: Continuous Variables Estimation Through Classification Networks Ensembles
• タイトル / 言語: en
• タイトル: Continuous Variables Estimation Through Classification Networks Ensembles
• 言語: eng
• 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, Nagoya University
• 著者所属: Graduate School of Information Science and Engineering, Ritsumeikan University
• 著者所属: Graduate School of Information Science and Engineering, Ritsumeikan University
• 著者所属(英): en
• 著者所属(英): Graduate School of Informatics, Nagoya University
• 著者所属(英): en
• 著者所属(英): Graduate School of Information Science and Engineering, Ritsumeikan University
• 著者所属(英): en
• 著者所属(英): Graduate School of Information Science and Engineering, Ritsumeikan University
• 著者名: Qianyuan, Liu; Yu, Wang; Jien, Kato
• 著者名(英): Qianyuan, Liu; Yu, Wang; Jien, Kato
• 論文抄録 / 内容記述タイプ: Other
• 論文抄録 / 内容記述: With the development of deep learning, CNNs have shown outstanding performance on various tasks. Previous approaches find that CNNs perform better on classification tasks than regression tasks, because regression task is a highly challenging task that approximates a mapping function from input variables to a continuous output variable. In the computer vision and multimedia communities, researchers address continuous variables estimation by deep convolutional neural regression networks. In this paper we make estimation of the continuous attributes of images by using classification networks ensembling. To the best of our knowledge, this is the first attempt to address regression problems through classification networks ensembles and our proposed method shows great versatility in different datasets. Experiment results show that our proposed method outperforms the regression method and single classification network.
• 論文抄録(英) / 内容記述タイプ: Other
• 論文抄録(英) / 内容記述: With the development of deep learning, CNNs have shown outstanding performance on various tasks. Previous approaches find that CNNs perform better on classification tasks than regression tasks, because regression task is a highly challenging task that approximates a mapping function from input variables to a continuous output variable. In the computer vision and multimedia communities, researchers address continuous variables estimation by deep convolutional neural regression networks. In this paper we make estimation of the continuous attributes of images by using classification networks ensembling. To the best of our knowledge, this is the first attempt to address regression problems through classification networks ensembles and our proposed method shows great versatility in different datasets. Experiment results show that our proposed method outperforms the regression method and single classification network.
• 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 2020-CVIM-221, 号 25, p. 1-6, 発行日 2020-03-09
• 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