學術產出-學位論文
文章檢視/開啟
書目匯出
-
題名 深度學習於國畫主題辨識之應用
Identifying Chinese painting genres with deep learning作者 許嘉宏
Hsu, Chia-Hung貢獻者 蔡炎龍
Tsai, Yen-Lung
許嘉宏
Hsu, Chia-Hung關鍵詞 深度學習
卷積神經網路
影像辨識
Deep Learning
Nerural Network
CNN
Image Recognition日期 2019 上傳時間 7-八月-2019 16:35:21 (UTC+8) 摘要 本篇文章主要使用卷積神經網路來進行圖像辨識，資料來源用台北故宮 博物院線上資料庫，其中圖像收藏量三萬筆，本篇將範圍縮小至畫軸的部 分，總計有 4 千筆，因為每張圖像有主要主題跟次要主題，無法直接用卷 積神經網路來分類。所以先利用 SLIC 演算法將圖像分割，再來進行標籤及 訓練模型。最後如有新的作品要進行辨識，也進行同樣分割，用模型辨識 後，再統整結果得到此作品有哪些主題性。
In this paper, we want to recognize one image with multiple genres. We collected data from National Palace Museun. If we just use traditional CNN to recognize it, we only get one genre with one image. Hence, we segment image with SLIC algorithm. It can segment image into fixed size with similar range, then we can use them to train the model. After training, if we get the new image, we can use SILC algorithm with same parameter and put it in the model. Then we can recognize this new image with multiple genres.參考文獻 [1]RadhakrishnaAchanta,AppuShaji,KevinSmith,AurelienLucchi,PascalFua,andSabine Süsstrunk. Slic superpixels, 2010.[2] Dzmitry Bahdanau, Kyunghyun Cho, and Yoshua Bengio. Neural machine translation by jointly learning to align and translate. CoRR, abs/1409.0473, 2014.[3] John Duchi, Elad Hazan, and Yoram Singer. Adaptive subgradient methods for online learning and stochastic optimization. J. Mach. Learn. Res., 12:2121–2159, 2011.[4] Ross B. Girshick, Jeff Donahue, Trevor Darrell, and Jitendra Malik. Rich feature hierarchies for accurate object detection and semantic segmentation. CoRR, abs/ 1311.2524, 2013.[5] J. B. Heaton, N. G. Polson, and J. H. Witte. Deep learning in finance. CoRR, abs/ 1602.06561, 2016.[6] Donald Hebb. The The Organization of Behavior. 1949.[7] Alex Krizhevsky, Ilya Sutskever, and Geoffrey E. Hinton. Imagenet classification with deep convolutional neural networks. In Peter L. Bartlett, Fernando C. N. Pereira, Christopher J. C. Burges, Léon Bottou, and Kilian Q. Weinberger, editors, NIPS, pages 1106–1114, 2012.[8] Alex Krizhevsky, Ilya Sutskever, and Geoffrey E Hinton. Imagenet classification with deep convolutional neural networks. In F. Pereira, C. J. C. Burges, L. Bottou, and K. Q. Weinberger, editors, Advances in Neural Information Processing Systems 25, pages 1097– 1105. Curran Associates, Inc., 2012.[9] Jan Kukačka, Vladimir Golkov, and Daniel Cremers. Regularization for deep learning: A taxonomy, 2017.[10]S.Lawrence,C.L.Giles,AhChungTsoi,andA.D.Back.Facerecognition:aconvolutional neural-network approach. Neural Networks, IEEE Transactions on, 8(1):98–113, January 1997.[11] Yann LeCun, Yoshua Bengio, and Geoffrey Hinton. Deep learning. Nature, 521(7553): 436–444, may 2015.[12] Min Lin, Qiang Chen, and Shuicheng Yan. Network in network, 2013.[13] National Palace Museun. 書畫典藏資料檢索系統, 2019.[14] Sinno Jialin Pan and Qiang Yang. A survey on transfer learning. IEEE Trans. on Knowl. and Data Eng., 22(10):1345–1359, October 2010.[15] Tara N. Sainath, Abdel rahman Mohamed, Brian Kingsbury, and Bhuvana Ramabhadran. Deep convolutional neural networks for lvcsr. In Acoustics, Speech and Signal Processing (ICASSP), 2013 IEEE International Conference on, 2013.[16] K. Simonyan and A. Zisserman. Very deep convolutional networks for large-scale image recognition. In International Conference on Learning Representations, 2015.[17] Jonathan J Tompson, Arjun Jain, Yann LeCun, and Christoph Bregler. Joint training of a convolutional network and a graphical model for human pose estimation. In Z. Ghahramani, M. Welling, C. Cortes, N. D. Lawrence, and K. Q. Weinberger, editors, Advances in Neural Information Processing Systems 27, pages 1799–1807. Curran Associates, Inc., 2014.[18] H. Y. Xiong, B. Alipanahi, L. J. Lee, H. Bretschneider, D. Merico, R. K. C. Yuen, Y. Hua, S. Gueroussov, H. S. Najafabadi, T. R. Hughes, Q. Morris, Y. Barash, A. R. Krainer, N. Jojic, S. W. Scherer, B. J. Blencowe, and B. J. Frey. The human splicing code reveals new insights into the genetic determinants of disease. Science, 347(6218):1254806– 1254806, dec 2014. 描述 碩士
國立政治大學
應用數學系
104751003資料來源 http://thesis.lib.nccu.edu.tw/record/#G0104751003 資料類型 thesis dc.contributor.advisor 蔡炎龍 zh_TW dc.contributor.advisor Tsai, Yen-Lung en_US dc.contributor.author (作者) 許嘉宏 zh_TW dc.contributor.author (作者) Hsu, Chia-Hung en_US dc.creator (作者) 許嘉宏 zh_TW dc.creator (作者) Hsu, Chia-Hung en_US dc.date (日期) 2019 en_US dc.date.accessioned 7-八月-2019 16:35:21 (UTC+8) - dc.date.available 7-八月-2019 16:35:21 (UTC+8) - dc.date.issued (上傳時間) 7-八月-2019 16:35:21 (UTC+8) - dc.identifier (其他 識別碼) G0104751003 en_US dc.identifier.uri (URI) http://nccur.lib.nccu.edu.tw/handle/140.119/124868 - dc.description (描述) 碩士 zh_TW dc.description (描述) 國立政治大學 zh_TW dc.description (描述) 應用數學系 zh_TW dc.description (描述) 104751003 zh_TW dc.description.abstract (摘要) 本篇文章主要使用卷積神經網路來進行圖像辨識，資料來源用台北故宮 博物院線上資料庫，其中圖像收藏量三萬筆，本篇將範圍縮小至畫軸的部 分，總計有 4 千筆，因為每張圖像有主要主題跟次要主題，無法直接用卷 積神經網路來分類。所以先利用 SLIC 演算法將圖像分割，再來進行標籤及 訓練模型。最後如有新的作品要進行辨識，也進行同樣分割，用模型辨識 後，再統整結果得到此作品有哪些主題性。 zh_TW dc.description.abstract (摘要) In this paper, we want to recognize one image with multiple genres. We collected data from National Palace Museun. If we just use traditional CNN to recognize it, we only get one genre with one image. Hence, we segment image with SLIC algorithm. It can segment image into fixed size with similar range, then we can use them to train the model. After training, if we get the new image, we can use SILC algorithm with same parameter and put it in the model. Then we can recognize this new image with multiple genres. en_US dc.description.tableofcontents 第一章 Introduction 1第二章 Deep Learning 3第一節 Neurons and Neural Networks 4第二節 Activation Function 7第三節 Loss Function 9第四節 Gradient Descent Method 11第三章 Convolutional Neural Network 13第一節 Convolution Layer 13第二節 Pooling Layer 22第四章 Data Collection and Processing 25第一節 K-means Clustering 25第二節 SLIC Algorithm 27第三節 Make the Labels 29第五章 Model Construction 33第一節 Models Structure 33第二節 Transfer Learning 37第三節 Imbalance Data 38第四節 Result 38第六章 Conclusion 41Appendix A Python Script 42A.1 Segment the Painting 42A.2 Create the Label 45A.3 Train the Model with the Model D 47A.4 Train the model with Transfer Learning 50A.5 GUI 53Bibliography 64 zh_TW dc.format.extent 13356256 bytes - dc.format.mimetype application/pdf - dc.source.uri (資料來源) http://thesis.lib.nccu.edu.tw/record/#G0104751003 en_US dc.subject (關鍵詞) 深度學習 zh_TW dc.subject (關鍵詞) 卷積神經網路 zh_TW dc.subject (關鍵詞) 影像辨識 zh_TW dc.subject (關鍵詞) Deep Learning en_US dc.subject (關鍵詞) Nerural Network en_US dc.subject (關鍵詞) CNN en_US dc.subject (關鍵詞) Image Recognition en_US dc.title (題名) 深度學習於國畫主題辨識之應用 zh_TW dc.title (題名) Identifying Chinese painting genres with deep learning en_US dc.type (資料類型) thesis en_US dc.relation.reference (參考文獻) [1]RadhakrishnaAchanta,AppuShaji,KevinSmith,AurelienLucchi,PascalFua,andSabine Süsstrunk. Slic superpixels, 2010.[2] Dzmitry Bahdanau, Kyunghyun Cho, and Yoshua Bengio. Neural machine translation by jointly learning to align and translate. CoRR, abs/1409.0473, 2014.[3] John Duchi, Elad Hazan, and Yoram Singer. Adaptive subgradient methods for online learning and stochastic optimization. J. Mach. Learn. Res., 12:2121–2159, 2011.[4] Ross B. Girshick, Jeff Donahue, Trevor Darrell, and Jitendra Malik. Rich feature hierarchies for accurate object detection and semantic segmentation. CoRR, abs/ 1311.2524, 2013.[5] J. B. Heaton, N. G. Polson, and J. H. Witte. Deep learning in finance. CoRR, abs/ 1602.06561, 2016.[6] Donald Hebb. The The Organization of Behavior. 1949.[7] Alex Krizhevsky, Ilya Sutskever, and Geoffrey E. Hinton. Imagenet classification with deep convolutional neural networks. In Peter L. Bartlett, Fernando C. N. Pereira, Christopher J. C. Burges, Léon Bottou, and Kilian Q. Weinberger, editors, NIPS, pages 1106–1114, 2012.[8] Alex Krizhevsky, Ilya Sutskever, and Geoffrey E Hinton. Imagenet classification with deep convolutional neural networks. In F. Pereira, C. J. C. Burges, L. Bottou, and K. Q. Weinberger, editors, Advances in Neural Information Processing Systems 25, pages 1097– 1105. Curran Associates, Inc., 2012.[9] Jan Kukačka, Vladimir Golkov, and Daniel Cremers. Regularization for deep learning: A taxonomy, 2017.[10]S.Lawrence,C.L.Giles,AhChungTsoi,andA.D.Back.Facerecognition:aconvolutional neural-network approach. Neural Networks, IEEE Transactions on, 8(1):98–113, January 1997.[11] Yann LeCun, Yoshua Bengio, and Geoffrey Hinton. Deep learning. Nature, 521(7553): 436–444, may 2015.[12] Min Lin, Qiang Chen, and Shuicheng Yan. Network in network, 2013.[13] National Palace Museun. 書畫典藏資料檢索系統, 2019.[14] Sinno Jialin Pan and Qiang Yang. A survey on transfer learning. IEEE Trans. on Knowl. and Data Eng., 22(10):1345–1359, October 2010.[15] Tara N. Sainath, Abdel rahman Mohamed, Brian Kingsbury, and Bhuvana Ramabhadran. Deep convolutional neural networks for lvcsr. In Acoustics, Speech and Signal Processing (ICASSP), 2013 IEEE International Conference on, 2013.[16] K. Simonyan and A. Zisserman. Very deep convolutional networks for large-scale image recognition. In International Conference on Learning Representations, 2015.[17] Jonathan J Tompson, Arjun Jain, Yann LeCun, and Christoph Bregler. Joint training of a convolutional network and a graphical model for human pose estimation. In Z. Ghahramani, M. Welling, C. Cortes, N. D. Lawrence, and K. Q. Weinberger, editors, Advances in Neural Information Processing Systems 27, pages 1799–1807. Curran Associates, Inc., 2014.[18] H. Y. Xiong, B. Alipanahi, L. J. Lee, H. Bretschneider, D. Merico, R. K. C. Yuen, Y. Hua, S. Gueroussov, H. S. Najafabadi, T. R. Hughes, Q. Morris, Y. Barash, A. R. Krainer, N. Jojic, S. W. Scherer, B. J. Blencowe, and B. J. Frey. The human splicing code reveals new insights into the genetic determinants of disease. Science, 347(6218):1254806– 1254806, dec 2014. zh_TW dc.identifier.doi (DOI) 10.6814/NCCU201900448 en_US
