近年來人工智慧的興起，使得人工智慧的應用不斷的出現在生活中(例如：交通、醫療、金融等），代表人工智慧需要具有像人類一樣擁有包括感知、學習、如何操作或移動事物。由於深度學習廣泛應用於人工智慧中，不論是圖像辨識、自然語言處理、手寫辨識均有顯著之成效。使深度學習有重大突破原因與其中的卷積神經網路密不可分，它是深度神經網路在影像辨識領域中發展的主要因素。
本論文將研究歷年ILSVRC比賽中突出之優勝隊伍方法，ILSVRC 2010錯誤率高達28%，在2012年AlexNet用卷積神經網路使錯誤率降低為15.3%，獲得ILSVRC 2012冠軍殊榮，使卷積神經網路技術成為眾多企業與學者研究之方向。在2014年GoogLeNet將電腦視覺的表現能夠與人類視覺相比。在2017年SeNet將錯誤率遠低於人類視覺的5%至10%，其錯誤率降低至2.25%。經過多年發展，深度神經網路中的隱藏層已經從兩層以上達到至數十層以上，訓練資料量類型、訓練時間、錯誤率將不同以往，於隱藏層內不同技術將使人工智慧更加接近人類思維。研究結果將顯示出各種基於深度學習之方法技術。

The rise of artificial intelligence in recent years. Make the application of artificial intelligence appear in life constantly (e.g. traffic, medical, commerce, etc.). Deep learning is applied to artificial intelligence that significant results have been achieved in image recognition, natural language processing, and handwriting recognition. The reason for the development of deep learning is related to the convolutional neural network. It is the main factor for the development of deep neural networks in image recognition.
In this paper, we study the every years in ILSVRC. Error rate of ILSVRC 2010 is 28%. In 2012, AlexNet uses a convolutional neural network reduced the error rate to 15.3%, Won the championship in ILSVRC 2012, convolutional neural network technology has become the research direction for enterprises and scholars. In 2014, GoogLeNet compared the performance of computer vision with human vision. In 2017, SeNet will have an error rate that is much lower than 5% to 10% of human vision, and its error rate is reduced to 2.25%. After years of development, the hidden layer in the deep neural network reaches more than tens of layers, the type of training data, training time, and error rate will be different. Different technologies within the hidden layer will bring artificial intelligence closer to human. The research results will show a variety of techniques using deep learning.

摘要........................i
Abstract...................ii
誌謝......................iii
目錄.......................iv
表目錄......................v
圖目錄.....................vi
符號說明..................vii
第一章 緒論................1
1.1 研究背景............1
1.2 論文架構............2
第二章 機器學習基礎.........3
2.1 卷積神經網路架構.....3
2.2 卷積層..............4
2.3 激活函數............5
2.4 池化層..............7
2.5 全連接層............8
第三章 相關技術............9
3.1 AlexNet............9
3.2 ZFNet.............11
3.3 GoogLeNet.........12
3.4 ResNet............16
3.5 GBD-Net...........18
3.6 SeNet.............20
第四章 實驗結果...........22
第五章 結論..............23
參考文獻...................24
Extended Abstract.........26

[ 1 ] Y. Lecun, L. Bottou, Y. Bengio, and P. Haffner, “Gradient-based learning applied to document recognition”, Proceedings of the IEEE, Vol.86, No.11, pp. 2278-2324.
[ 2 ] G. Cybenko, “Approximation by superpositions of a sigmoidal function”, Mathematics of Control, Signals and Systems, New York, USA, 1989, pp. 303-314.
[ 3 ] K. Jarrett, K. Kavukcuoglu, M. Ranzato, and Y. LeCun, ”What is the best multi-stage architecture for object recognition?”, IEEE 12th International Conference on Computer Vision, Kyoto, Japan, 2009, pp. 2146-2153.
[ 4 ] V. Nair and G. E. Hinton, “Rectified linear units improve restricted boltzmann machines”, 27th International Conference on Machine Learning, Haifa, Israel, 2010.
[ 5 ] A. Krizhevsky, I. Sutskever, G. E. Hinton, “ImageNet Classification with Deep Convolutional Neural Networks”, Neural Information Processing Systems(NIPS), Lake Tahoe, Nevada, USA, 2012.
[ 6 ]Y. L. Cun, B. Boser, J.S. Denker, R.E. Howard, W. Habbard, L.D. Jackel, and D. Henderson, “Handwritten digit recognition with a back-propagation network”, Neural Information Processing Systems(NIPS), San Francisco, USA, 1989.
[ 7 ] D. Ciregan, U. Meier, and J. Schmidhuber, ”Multi-column deep neural networks for image classification”, IEEE Conference on Computer Vision and Pattern Recognition, Providence, USA, 2012, pp. 3642-3649.
[ 8 ] G.E. Hinton, N. Srivastava, A. Krizhevsky, I. Sutskever, and R.R. Salakhutdinov, “Improving neural networks by preventing co-adaptation of feature detectors”, arXiv preprint arXiv:1207.0580, 2012.
[ 9 ] M. D. Zeiler and Rob Fergus, “Visualizing and Understanding Convolutional Networks”, European Conference on Computer Vision, Zurich, Switzerland, 2014, pp. 818-833
[ 10 ] M. D. Zeiler, G. W. Taylor, and R. Fergus, “Adaptive deconvolutional networks for mid and high level feature learning”, International Conference on Computer Vision, Barcelona, Spain, 2011, pp. 2018-2025.
[ 11 ] C. Szegedy, W. Liu, Y. Jia, P. Sermanet, S. Reed, D. Anguelov, D. Erhan, V. Vanhoucke, and A. Rabinovich, “Going deeper with convolutions”, IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Boston, USA, 2015.
[ 12 ] M. Lin, Q. Chen, and S. Yan, “Network In Network”, International Conference on Learning Representations (ICLR), Banff, Canada, 2014.
[ 13 ] K. He, X. Zhang, S. Ren, and J. Sun, ”Deep Residual Learning for Image Recognition”, IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, USA, 2016, pp. 770-778.
[ 14 ] X. Zeng, W. Ouyang, J. Yan, H. Li, T. Xiao, K. Wang, Y. Liu, Y. Zhou, B. Yang, Z. Wang, H. Zhou, and X. Wang, “Crafting GBD-Net for Object Detection”, arXiv preprint arXiv:1610.02579, 2016.
[ 15 ] S. Gidaris, N. Komodakis, “Object Detection via a Multi-region and Semantic Segmentation-Aware CNN Model”, IEEE International Conference on Computer Vision (ICCV), Santiago, Chile, 2015, pp. 1134-1142.
[ 16 ] J. Hu, L. Shen, and G. Sun, ” Squeeze-and-Excitation Networks”, IEEE/CVF Conference on Computer Vision and Pattern Recognition, Salt Lake City, USA, 2018, pp. 7132-7141.