Abstract: Scene Classification is a subdivision problem of Large-scale classfication problem since the latter has been basically resolved. In this article, several common Scene Classification Data-set and their differences are introduced. Additionally, there are lots of advanced methods of Deep Convolutional Neural Network. Methods for solving Large-scale Classification problems to be used on solving Scene Classification is a very common way. This article summerizes the results of those network structures trained on Scene Data-sets. Therefore, this article introduces some improvement for simply using CNN on Scene Classification and their better result. Since the common network structure is so complicated that it takes a long time to train and test, a method of simplifying these deep networks is raised in this article. Reducing size of input pictures and numbers of convolution kernels could take effect on increasing the speed on both training and testing stages. Finally, this much smaller network got an acceptable result on the data-set. % Reviews: please describe the background, status and application of the research with 150-300 words. I and we can not be used as the subject, % and the abstract must not the same as the sentences of the main text. General research paper: please extracts the key points of the paper, give the main research achievements with object, methods, results and conclusion with 200-400 words. I and we can not be used as the subject, and the abstract must not the same as the sentences of the main text.

Alert Name:
Alerting to:
Authentication email will be sent to your email address in 24 hours
Frequency:
Email Message Format:	Plain text Graphical(HTML)

Complete the form below and we will recommend the selected titles to your friends on your behalf. * Indicates a required field.
Your name*:
Your email address*:
Recipient's name*:
Recipient's email address*:
(multiple recipient's names and email addresses should be separated with semicolons)
Your comments:	I thought you would find the page(s) useful.

Your name:
Your email address:
Recipient's name:
Recipient's email address:
(multiple recipient's names and email addresses should be separated with semicolons)
Your comments:	I thought you would find this page useful.

Disclaimer: This message was sent to your friend using the "Send it to a friend" facility on the Sciencepaper Online’ WWW site, http://www.paper.edu.cn/en. The Sciencepaper Online is not responsible for the content of this email, and anything said in this email does not necessarily reflect the Sciencepaper Online's views.


	1. Scene Classification Based on minimized Deep Convolutional Neural Networks
	LIU Yu-xuan, DONG Yuan, BAI Hong-liang
	Electrics, Communication and Autocontrol Technology 24 June 2016
	Show/Hide Abstract \| Cite this paper︱Full-text: PDF (0 B)

	Abstract: Scene Classification is a subdivision problem of Large-scale classfication problem since the latter has been basically resolved. In this article, several common Scene Classification Data-set and their differences are introduced. Additionally, there are lots of advanced methods of Deep Convolutional Neural Network. Methods for solving Large-scale Classification problems to be used on solving Scene Classification is a very common way. This article summerizes the results of those network structures trained on Scene Data-sets. Therefore, this article introduces some improvement for simply using CNN on Scene Classification and their better result. Since the common network structure is so complicated that it takes a long time to train and test, a method of simplifying these deep networks is raised in this article. Reducing size of input pictures and numbers of convolution kernels could take effect on increasing the speed on both training and testing stages. Finally, this much smaller network got an acceptable result on the data-set. % Reviews: please describe the background, status and application of the research with 150-300 words. I and we can not be used as the subject, % and the abstract must not the same as the sentences of the main text. General research paper: please extracts the key points of the paper, give the main research achievements with object, methods, results and conclusion with 200-400 words. I and we can not be used as the subject, and the abstract must not the same as the sentences of the main text.
	TO cite this article:LIU Yu-xuan, DONG Yuan, BAI Hong-liang. Scene Classification Based on minimized Deep Convolutional Neural Networks[OL].[24 June 2016] http://en.paper.edu.cn/en_releasepaper/content/4698285


	2. Masked Face Detection Using Deep Learning
	Li yichao,Qiting Ye,Zhao Luo,Shiming Ge
	Electrics, Communication and Autocontrol Technology 21 March 2016
	Show/Hide Abstract \| Cite this paper︱Full-text: PDF (0 B)

	Abstract:Face occlusion such as masked face is a challenge problem for most face detection algorithms due to a lack of discriminative information. In this paper, we proposed a novel method to address occluded face detection especially masked face detection. We built a masked face database whose images are collected from web images in the wild. The database includes more than 6000 images and more than 10000 masked faces. To perform masked face detection, we proposed a joint pre-detection and classification method, which learn a discriminative classifier based on deep learning to classify the face proposals which are generated by some weak face detectors. The classifier has higher discrimination power to masked face, unmasked face and non-face. Experimental comparisons with state-of-the-art face detection methods show that the proposed method can give better performance. .
	TO cite this article:Li yichao,Qiting Ye,Zhao Luo, et al. Masked Face Detection Using Deep Learning[OL].[21 March 2016] http://en.paper.edu.cn/en_releasepaper/content/4681179

Saved Papers

Saved Papers


	3. Learning-Based Compressed Sensing for Infrared Image Super Resolution
	Yao Zhao,Xiubao Sui,Qian Chen,Shaochi Wu
	Electrics, Communication and Autocontrol Technology 30 November 2015
	Show/Hide Abstract \| Cite this paper︱Full-text: PDF (0 B)

	Abstract:This paper presents an infrared image super-resolution method based on compressed sensing (CS). First, the reconstruction model under the CS framework is established and a Toeplitz matrix is selected as the sensing matrix. Compared with traditional learning-based methods, the proposed method uses a set of sub-dictionaries instead of two coupled dictionaries to recover high resolution (HR) images. And Toeplitz sensing matrix allows the proposed method time-efficient. Second, all training samples are divided into several feature spaces by using the proposed adaptive k-means classification method, which is more accurate than the standard k-means method. On the basis of this approach, a complex nonlinear mapping from the HR space to low resolution (LR) space can be converted into several compact linear mappings. Finally, the relationships between HR and LR image patches can be obtained by multi-sub-dictionaries and HR infrared images are reconstructed by the input LR images and multi-sub-dictionaries. The experimental results show that the proposed method is quantitatively and qualitatively more effective than other state-of-the-art methods.
	TO cite this article:Yao Zhao,Xiubao Sui,Qian Chen, et al. Learning-Based Compressed Sensing for Infrared Image Super Resolution[OL].[30 November 2015] http://en.paper.edu.cn/en_releasepaper/content/4665509


	4. The Research and Improvement on Dark Channel Prior Image Dehazing Algorithm
	Han Wang,Bo Yang
	Electrics, Communication and Autocontrol Technology 18 October 2012
	Show/Hide Abstract \| Cite this paper︱Full-text: PDF (0 B)

	Abstract:This paper focuses on the improvement of a newly proposed single image dehazing algorithm, Dark Channel Prior Method which is firstly proposed in paper Single Image Haze Removal Using Dark Channel Prior. The dark channel prior dehazing algorithm is based on a significant observation that in haze-free outdoor images most regions contain some pixels which have very low intensities (dark channel) in at least one color channel (RGB). Based on this observation, we can use the dark channel as the prior to directly estimate the haze thickness and recover a high quality haze-free image. Compared with other methods for single image dehazing, dark channel prior not only has better performance in situations of dense haze, but also do not rely much on significant variance on transmission and color information in the input image. Based on that, the dark channnel prior method, as a novel haze removing method, provides a simpler and more effective way for single image haze removal. Apart from that, an improvement of the original dark channel dehazing algorithm, mainly focus on removing the noise artefacts showed in the dehazed images, will be paid more attention. Through the comparison between the original and refined dark channel prior dehazing algorithm, we can see that by using the dynamic estimation algorithm on the value of t0, which is proposed in this paper, can significantly improve the dehazing performance and the dehazed image quality of the original dark channel prior dehazing algorithm.
	TO cite this article:Han Wang,Bo Yang. The Research and Improvement on Dark Channel Prior Image Dehazing Algorithm[OL].[18 October 2012] http://en.paper.edu.cn/en_releasepaper/content/4491711


	5. An Effectively License Plate Localization Method
	Dai Yourui,ZhangHonggang,LiQiaohong
	Electrics, Communication and Autocontrol Technology 15 December 2011
	Show/Hide Abstract \| Cite this paper︱Full-text: PDF (0 B)

	Abstract:Benefit from the lower cost and the easier installation, Intelligent Traffic system (ITS) based on high definition (HD) picture is increasingly used. In this paper, we describe a edge and color feature based method, which is able to locate and extract license plate from the high definition images in real-time. In the proposed algorithm, the edge feature is fully utilized to find the candidate region of the license plate (LP). And then, a edge identity similarity method is employed to adjust the position of each candidate, which can get the accurate LP position. Finally, an effective evaluation method is proposed to get the real LP from the candidate region. This thoroughly integrated system can obtain accurate localization of LP real-time in the HD traffic pictures under complicated background and changing illumination conditions. The experiment results indicate that the presented system is excellent in accuracy with a good speed.
	TO cite this article:Dai Yourui,ZhangHonggang,LiQiaohong. An Effectively License Plate Localization Method[OL].[15 December 2011] http://en.paper.edu.cn/en_releasepaper/content/4453850


	6. Design And Implementation of Digital Cinema Security System
	Peng Liyang
	Electrics, Communication and Autocontrol Technology 18 October 2010
	Show/Hide Abstract \| Cite this paper︱Full-text: PDF (0 B)

	Abstract:Thanks to its advantages that beat the traditional film, digital cinema is becoming more and more popular. But its strong points also bring some problems, one of which is that its convenience of storage and transmission makes it more likely to be pirated. Therefore, security is very important to digital cinema. This paper designs a security system which includes digital certificate manager, digital cinema issuer, encrypted digital cinema package producer and digital cinema manager, based on some security key techniques and digital cinema standards. The paper will clarify the functions of each part and explain how they work together. In order to be practical, the system is designed to be compliant with the specifications released by Digital Cinema Initiatives, LLC (DCI).
	TO cite this article:Peng Liyang. Design And Implementation of Digital Cinema Security System[OL].[18 October 2010] http://en.paper.edu.cn/en_releasepaper/content/4388628


	7. New Adaptive Demosaicking Methods Based on Color Difference for Bayer Color Filter Array
	Wang Hongwei ,Cheng Qongqiang
	Electrics, Communication and Autocontrol Technology 07 April 2009
	Show/Hide Abstract \| Cite this paper︱Full-text: PDF (0 B)

	Abstract:Single-chip digital cameras use color filter arrays (CFA) to sample different color component. So demosaicking algorithms to complete RGB values is necessary. In this paper, we introduce the data adaptive filtering concept, exploit it for demosaicking, and finally propose a novel demosaicking algorithm for Bayer color filter array. Using a different function mapping, we adopt the color difference model, predict the color difference value at the R/B location and then update it. Through the refined color difference value, we first interpolate the full G channel, and then recover R and B channels. The efficiency of the proposed approach is demonstrated by experimental results with simulated and real data. It not only performs the high quality of our approach visually but also has a notably lower computational cost.
	TO cite this article:Wang Hongwei ,Cheng Qongqiang . New Adaptive Demosaicking Methods Based on Color Difference for Bayer Color Filter Array[OL].[ 7 April 2009] http://en.paper.edu.cn/en_releasepaper/content/31103


	8. Pedestrian Detection by Boosting Neural Networks
	Jia Hui-Xing,Zhang Yu-Jin
	Electrics, Communication and Autocontrol Technology 17 December 2008
	Show/Hide Abstract \| Cite this paper︱Full-text: PDF (0 B)

	Abstract:In this paper, a fast pedestrian detection system by boosting neural network classifiers is built. The object to be detected is represented by a collection of blocks. For each block, the histogram of orientated gradients feature is extracted and a neural network classifier is built as weak hypothesis. Then these hypotheses are selected sequentially by Gentle Adaboost, and the cascade structure is used to speedup the detector. Compared to global linear SVM classifiers, the new method gets better performance on the INRIA pedestrian detection database at a much faster speed.
	TO cite this article:Jia Hui-Xing,Zhang Yu-Jin. Pedestrian Detection by Boosting Neural Networks[OL].[17 December 2008] http://en.paper.edu.cn/en_releasepaper/content/26680

	Check out RSS, or use RSS reader to subscribe this item