检索结果-内蒙古大学图书馆

16th International Forum on Digital TV and Wireless Multimedia communication, IFTC 2019

作者： Zheng, Chuanyang Zhou, Jun Sun, Jun Zhao, Lihua Institute of Image Communication and Network Engineering Shanghai Jiao Tong University Shanghai200240 China Shanghai Key Lab of Digital Media Processing and Transmissions Shanghai Jiao Tong University Shanghai200240 China Children’s Hospital of Shanghai Shanghai200062 China

ISBN: (纸本)9789811533402

Non-invasive gaze estimation from only eye images captured by camera is a challenging problem due to various eye shapes, eye structures and image qualities. Recently, CNN network has been applied to directly regress eye image to gaze direction and obtains good performance. However, generic approaches are susceptible to bias and variance highly relating to different individuals. In this paper, we study the person-specific bias when applying generic methods on new person. And we introduce a novel appearance-based deep neural network integrating meta-learning to reduce the person-specific bias. Given only a few person-specific calibration images collected in normal calibration process, our model adapts quickly to test person and predicts more accurate gaze directions. Experiments on public MPIIGaze dataset and Eyediap dataset show our approach has achieved competitive accuracy to current state-of-the-art methods and are able to alleviate person-specific bias problem. © 2020, Springer Nature Singapore Pte Ltd.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Motion compensation optimization of H.264 decoder based on DSP hardware platform

引用

Nanjing Youdian Daxue Xuebao (Ziran Kexue Ban)/Journal of Nanjing University of Posts and Telecommunications (Natural Science) 2007年第2期27卷 80-84页

作者： Gan, Zong-Liang Wang, Kai Zhu, Xiu-Chang Jiangsu Province Key Lab on Image Processing and Image Communication Nanjing University of Posts and Telecommunications Nanjing 210003 China

The H.264 video coding standard has a significant performance benefit versus former standards. However, motion compensation decoding process takes up a significant decoding time due to the advanced coding features. In this paper, the theory of motion compensation decoding in H.264 is given at first. Then the improved method is proposed by motion compensation fast decoding algorithm and optimization based on ADI Blackfin533 hardware platform. Both PC and ADI Blackfin533 experimental results show that the proposed optimization strategy can substantially improve the efficiency of real-time H.264 decoder.

关键词： Digital signal processor Direct memory access H.264 Motion compensation decoding

来源：评论

学校读者我要写书评

暂无评论

Anchored Neighborhoods Search Based on Global Dictionary Atoms for Face Photo-Sketch Synthesis

Anchored Neighborhoods Search Based on Global Dictionary Ato...

引用

2019第十一届数字图像处理国际会议

作者： Feng Liu Ran Xu Jieying Zheng Qiuli Lin Zongliang Gan College of Educational Science and Technology Nanjing University of Posts and Telecommunications Jiangsu Province Key Lab on Image Processing & Image Communications Nanjing University of Posts and Telecommunications Key Lab of Broadband Wireless Communication and Sensor Network Technology Ministry of Education Nanjing University of Posts and Telecommunications

Example-based face sketch synthesis technology generally requires face photo-sketch images with face alignment and size *** break through the limitation,we propose a global face sketch synthesis method:In training,all training photo-sketch patch pairs are collected together and a photo feature dictionary is learned from the photo *** each atom of the dictionary,its K closest photo-sketch patch pairs are clustered,namely "Anchored Neighborhood".In testing,for each test photo patch,we search its nearest photo patch in the Anchored Neighborhood determined by its closest atom,then the corresponding sketch patch is the *** the same way,we train and test in the high-frequency domain and synthesis the high-frequency ***,the fusion of the initial and the high-frequency results is the final *** experiments on three public face sketch datasets and various real-world photos demonstrate the effectiveness and robustness of the proposed method.

关键词： Face sketch synthesis Dictionary learning Dictionary atom Anchored neighborhood image patch

来源：评论

学校读者我要写书评

暂无评论

FC-GNN: Recovering Reliable and Accurate Correspondences from Interferences

FC-GNN: Recovering Reliable and Accurate Correspondences fro...

引用

Conference on Computer Vision and Pattern Recognition (CVPR)

作者： Haobo Xu Jun Zhou Hua Yang Renjie Pan Cunyan Li Institute of Image Communication and Network Engineering Shanghai Jiao Tong University Shanghai Key Lab of Digital Media Processing and Transmission

ISBN: (数字)9798350353006

ISBN: (纸本)9798350353013

Finding correspondences between images is essential for many computer vision tasks and sparse matching pipelines have been popular for decades. However, matching noise within and between images, along with inconsistent key-point detection, frequently degrades the matching performance. We review these problems and thus propose: 1) a novel and unified Filtering and Calibrating (FC) approach that jointly rejects outliers and optimizes inliers, and 2) leveraging both the matching context and the underlying image texture to remove matching uncertainties. Under the guidance of the above innovations, we construct Filtering and Calibrating Graph Neural Network (FC-GNN), which follows the FC approach to recover reliable and accurate correspondences from various interferences. FC-GNN conducts an effectively combined inference of contextual and local information through careful embedding and multiple information aggregations, predicting confidence scores and calibration offsets for the input correspondences to jointly filter out outliers and improve pixel-level matching accuracy. Moreover, we exploit the local coherence of matches to perform inference on local graphs, thereby reducing computational complexity. Overall, FC-GNN operates at lightning speed and can greatly boost the performance of diverse matching pipelines across various tasks, showcasing the immense potential of such approaches to become standard and pivotal components of image matching. Code is avaiable at https://***/xuy123456/fcgnn.

关键词： Matched filters Computer vision Technological innovation Accuracy Uncertainty Computer network reliability Pipelines

来源：评论

学校读者我要写书评

暂无评论

Vehicle Color Recognition Based on Superpixel Features

Vehicle Color Recognition Based on Superpixel Features

引用

2019第十一届数字图像处理国际会议

作者： Qiuli Lin Feng Liu Qiang Zhao Ran Xu Jiangsu Province Key Lab on Image Processing & Image Communications Nanjing University of Posts and Telecommunications College of Educational Science and Technology Nanjing University of Posts and Telecommunications Key Lab of Broadband Wireless Communication and Sensor Network Technology Ministry of Education Nanjing University of Posts and Telecommunications

In this paper,a novel methodology is presented to settle the region of interest(ROI) detection problem in vehicle color recognition so as to remove the redundant components of vehicles that interfere greatly with color *** order to make full use of the local color and spatial information,vehicle images are divided into different superpixels at *** spatial relationship between superpixels and the outermost pixels is then used for the background removal of vehicle *** comparing with the vehicle window clustering centroids obtained by k-means,the superpixels close to the universal color characteristics of windows are removed so that the dominant color superpixels are ***,a linear Support Vector Machine classifier is trained for color *** experiments demonstrate that the proposed methodology is effective for color region of interest detection and thus contribute to vehicle color recognition.

关键词： vehicle color recognition region of interest superpixel universal color characteristics Support Vector Machine

来源：评论

学校读者我要写书评

暂无评论

Effective Scheme for Global Abnormal Event Detection for Surveillance Video

Effective Scheme for Global Abnormal Event Detection for Sur...

引用

2016 International Conference on communications, Information Management and Network Security (CIMNS2016)

作者： Fangxu Dong Dong Hu Education Ministry's Key Lab of Broadband Wireless Communication and Sensor Network Technology Education Ministry's Engineering Research Center of Ubiquitous Network and Heath Service Jiangsu Province's Key Lab of Image Processing and Image Communications Nanjing University of Posts and Telecommunications

ISBN: (纸本)9781510830981

An effective algorithm for global abnormal detection from surveillance video is proposed in this paper. The algorithm is based on sparse representation. To deal with the illumination change in video scenes, specific feature extract methods are designed for corresponding illumination conditions. In the case of non-uniform illumination, features are extracted directly on the original image;in the case of uniform illumination, features are extracted on the binary image obtained by threshold segmentation on the difference image, where the thresholds are computed by the Otsu's method. The features extracted on normal video are used to learn an over-complete dictionary. Then, the sparse reconstruction cost over the dictionary is used to detect abnormal events. Experiments on the open global abnormal dataset and the comparison to the state-of-the-art methods validate effectiveness and quickness of our algorithm.

关键词： abnormal detection illumination change binary images sparse reconstruction cost

来源：评论

学校读者我要写书评

暂无评论

Fast encoding method using CU depth for quality scalable HEVC

Fast encoding method using CU depth for quality scalable HEV...

引用

IEEE Workshop on Advanced Research and Technology in Industry Applications (WARTIA)

作者： Qingyang Ge Dong Hu Education Ministry's Key Lab of Broadband Wireless Communication and Sensor Network Technology Jiangsu Province's Key Lab of Image Procession and Image Communications Nanjing University of Posts and Telecommunications Nanjing CN

In this paper, a new fast encoding scheme using CU depth information for quality scalable HEVC was proposed. The proposed scheme based on the correlation between the base layer and enhancement layer, as well as the correlation between current frame and reference frame, which obtained by experiment. The main idea of the algorithm is that when each LCU is encoded in enhancement layer, it's corresponding encoded LCU's best coding depth in base layer is used as the new maximum encoding depth in enhancement layer, thereby, it can reduce the unnecessary CU division process then reduce the computational complexity. On the other hand, for base layer, when each LCU encoding in current frame, the corresponding LCU's depth in already coded reference frame is utilized to skip parts process of mode selection before reach this reference depth, so that the computational complexity can be reduced again. Experimental results show that compared with the original SHM5.1, our proposed method saves about 59.45% and 67.71% encoding time in whole encoding layers and enhancement layer respectively, with a negligible decrease in PSNR of 0.014 dB and bitrate increase of 0.46% in EL, as well as PSNR of 0.05 dB and bitrate increase of 1.15% in BL averagely.

关键词： Encoding Scalability Computational complexity Correlation PSNR Bit rate

来源：评论

学校读者我要写书评

暂无评论

A new inter prediction mode decision method based on temporal correlation for HEVC encoder

A new inter prediction mode decision method based on tempora...

引用

IEEE Workshop on Advanced Research and Technology in Industry Applications (WARTIA)

作者： Qiuping Xu Dong Hu Education Ministry's Key Lab of Broadband Wireless Communication and Sensor Network Technology Jiangsu Province's Key Lab of Image Procession and Image Communications Nanjing University of Posts and Telecommunications Nanjing CN

To alleviate the encoder computation load in inter-frame coding of the High Efficiency Video Coding (HEVC), a new fast inter-frame prediction mode decision method is proposed by reducing the number of execution of the advanced motion vector prediction (AMVP) mode. Based on the correlation analysis between the corresponding located CU (Coding Unit) in the previous frame and the current CU, we first select the best one inter prediction mode by just performing merge mode decision process and then only perform the AMVP process on the selected one. Experimental results show that the proposed scheme provides 33.04% encoder time savings in Random Access (RA) configuration and 38.05% time savings in Low Delay (LD) configuration on average, and its increment in bit rate is 0.87% in RA and 1.29% in LD compared to the default encoding scheme in HM10.0.

关键词： Encoding Correlation Video coding Vectors Bit rate Algorithm design and analysis Copper

来源：评论

学校读者我要写书评

暂无评论

Disparity estimation algorithm for stereo video coding based on edge detecetion

Disparity estimation algorithm for stereo video coding based...

引用

International Conference on Wireless communications and Signal processing (WCSP)

作者： Jingjing Fan Feng Liu Weibin Bao Hongfei Xia Image Processing and Image Communication Jiangsu Key Laboratory Nanjing University of Posts and Telecommunications Nanjing China Zhijiang College Zhejiang University of Science and Technology Hangzhou China ZTE Telecom Company Limited Nanjing China

Disparity estimation is an important technique in stereo video coding. This paper presents a disparity estimation algorithm based on edge detection. The algorithm makes full use of the human visual characteristics, that is, the human eye is more sensitive to the distortion of the edge region. Therefore, joint estimation is used for edge detection. The large code block size for coding the background region and the flat areas while small size for coding the edge region were used in this paper. Compared to the disparity estimation algorithm proposed in, the proposed algorithm can greatly improve the encoding speed of stereo video without affecting subjective image quality.

关键词： Video coding image coding image edge detection Humans Motion estimation Educational institutions Scalability Redundancy image processing image communication

来源：评论

学校读者我要写书评

暂无评论

Group detection in complex scene by split-merge strategy

引用

ICIC Express Letters, Part B: Applications 2015年第12期6卷 3219-3224页

作者： Chen, Qiao Chen, Changhong Xie, Xiaofang Liu, Feng Key Lab of Image Processing and Image Communications of Jiangsu Province Nanjing University of Posts and Telecommunications No. 66 New Mofan Road Nanjing210003 China Department of Electrical Engineering Laiwu Vocational and Technical College Laiwu271100 China

Detection of groups of interacting people is a difficult task, especially in an unconstrained and crowded environment. In this paper, as a main contribution, we present a novel and efficient framework for social group detection, which is based on the theories of social distance model and social force model. Besides, we propose a new method to find the relationships between individuals in a scene, which is based on splitmerge strategy. The results demonstrate that our approach obtains effective performance in a more generic scene where many other methods fail. © 2015, ICIC Express Letters Office. All rights reserved.

关键词： Group theory

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：