Deep Learning Literature Reading notes (3)

21. Application of Depth neural network in visual significance (visual Attention with deep neural Networks) (English, conference papers, 2015, IEEE Search)

This article focuses on the application of CNN in the field of significance detection.

　　

22. Progress in deep learning Research (Chinese, Journal, 2015, net)

A summary article on deep learning is a detailed description of the origin of deep learning, the visual mechanism of human brain and the structure of CNN, and introduces the main improvement direction of deep learning in the future.

23. Progress in deep learning Research (Chinese, Journal, 2014, net)

It is important to emphasize that the groundbreaking work done by Hinton and others is mainly in the DBN, not CNN. This article mainly introduces the confidence network and the automatic Encoder, summarizes the improvement progress of initialization method, network layer, activation function selection, model structure, learning algorithm, and introduces the practical application of deep learning and the future research direction, is a more comprehensive review article.

24. Research and development of deep learning (Chinese, periodicals, 2015, net)

It is pointed out in the paper that the BP algorithm was introduced in 1974, and the principle and structure of limited Boltzmann machine, deep confidence network, automatic encoder and convolutional Neural network are presented, which has certain reference value for CNN introduction. At the same time, the application and research progress of deep learning are briefly introduced.

25. A summary of deep learning studies (Chinese, periodicals, 2015, net)

The introduction of CNN in detail, referring to the average pooling, maximum pooling and other noun concepts, a brief introduction of the Deepid and DeepFace algorithm, reference literature is relatively classic.

26. A summary of deep learning studies (Chinese, periodicals, 2012, net)

It is pointed out that LeCun and others first put forward CNN as a multi-layer structure learning algorithm, introduced the principle structure of DBN and CNN, in which the introduction of CNN is more detailed, and the application progress of both is briefly introduced.

27. Deep learning and its new progress in target and behavior recognition (Chinese, journal, 2014, net)

Written by the Tsinghua Daniel, is a typical review article in deep learning, which mentions the support of related neuroscience. This article mainly introduces the principle, structure and improvement scheme of the restricted Boltzmann machine and automatic encoder, and gives a detailed introduction to its application in object and behavior recognition, as well as the problems that need to be solved in the future, the article has a great reference value in both content and structure.

28. Overview of Image Object classification and detection algorithms (Chinese, periodicals, 2013, net)

The article summarizes today's classification detection algorithms (including deep learning) with visual contests such as the Pascal VOC contest and the Imagenet competition, providing detailed game data and convincing.

29. Light-invariant face recognition using convolutional neural networks (illumination invariant faces recognition using convolutional neural Networks) (English, conference papers, 2015, IEEE search)

The most basic application of CNN in the face recognition problem, the old method solves the new problem, the reason is the new problem, because the illumination uneven interference is added in the face recognition.

30. Age estimates based on deep feature learning (deep-learned Feature for Ages estimation) (English, conference papers, 2015, IEEE Search)

The first application of CNN in age estimation, mainly has the following two innovative points: first, the characteristics of the CNN mapping layer map are used, rather than just the top-level output feature map, is a multi-layered feature fusion idea, the second is to combine the manifold learning classifier with CNN, is a combination of CNN and advanced classifiers.

Copyright NOTICE: This article for Bo Master original article, without Bo Master permission not reproduced.

Deep Learning Literature Reading notes (3)