检索结果-内蒙古大学图书馆

IEEE International Symposium on Circuits and Systems

作者： Xiongkuo Min Guangtao Zhai Zhongpai Gao Ke Gu Institute of Image Communication and Information Processing Shanghai Jiao Tong University Shanghai China Shanghai Key Laboratory of Digital Media Processing and Transmissions

ISBN: (纸本)9781479934331

images usually contain areas that particularly attract people's attention and visual attention is an important feature of human visual system (HVS). Visual attention had been shown to be effective in improving performance of existing image quality assessment (IQA) metrics. However, with the quick advancement of IQA research, the booming of open IQA databases calls for associated comprehensive and accurate visual attention dataset. Despite of the large number of existing computational attention/saliency models, the most accurate measure of human attention is still human based. In this research, we first conduct extensive eye tracking experiments for all the pristine images from the seven widely used IQA databases (LIVE, TID2008, CSIQ, Toyama, LIVE Multiply Distortion, IVC and A57 databases). Then we propose a gaze-duration adaptive weighting approach to generate saliency maps from the eye tracking data. When applied on the IQA databases, experimental results suggest that accuracy of benchmark quality metrics, e.g. PSNR and SSIM can be systematically improved, outperforming existing saliency datasets. Both the eye tracking data and the saliency maps in this research will be made publicly available at ***.

关键词： important HVS IQA

来源：评论

学校读者我要写书评

暂无评论

THz Spectrum Deconvolution with Total Variation Regularization

THz Spectrum Deconvolution with Total Variation Regularizati...

引用

International Congress on image and Signal processing

作者： Lizhen Deng Hu Zhu Guanming Lu State Key Laboratory of Advanced Electromagnetic Engineering and Technology Huazhong University of Science and Technology Wuhan 430074 China Jiangsu Province Key Lab on Image Processing & Image Communication Nanjing University of Posts and Telecommunications Nanjing 210003 China

ISBN: (纸本)9781479927654

Deconvolution has become one of the most used methods for improving spectral resolution, and blind deconvolution as a typical method has been researched widely. However, the predefined point spread function (PSF) used in blind deconvolution method is not known exactly in practice. In general, the PSF is estimated simultaneously from the observed spectrum, but it becomes difficult when the spectroscopic data are polluted by strong noise. In this paper, we present a deconvolution method used to improve the resolution of THz spectrum. In the method, the energy function is constructed, which includes the likelihood term, Total variation of spectrum term and L2 norm of the PSF term. The PSF is modeled as a parametric function combination with the a priori knowledge about the characteristics of the instrumental response. The spectrum and the parameter of PSF are obtained by minimizing the energy functional using alternate minimization approach. Experimental results are shown to demonstrate the efficiency of the proposed method used for THz spectrum.

关键词： THz spectrum Total variation deconvolution alternate minimization approach

来源：评论

学校读者我要写书评

暂无评论

Error concealment for stereoscopic images using mode selection

Error concealment for stereoscopic images using mode selecti...

引用

IEEE International Symposium on Intelligent Signal processing and communication Systems (ISPACS)

作者： Lijuan Song Xiuchang Zhu State Grid Electric Power Research Institute Nanjing China Jiangsu Province Key Laboratory on Image Processing & Image Communication Nanjing University of Posts and Telecommunications Nanjing China

In this paper, a novel error concealment (EC) method for compressed stereoscopic image pairs is presented, which contains a new binocular EC mode and an improved monocular EC mode. The proposed algorithm selects appropriate EC mode to conceal the error block (EB) in the stereoscopic image according to the local characteristic of the EB. The experimental results demonstrate that the proposed scheme has good subjective and objective EC performance in stereoscopic images as compared to monocular mode.

关键词： PSNR Particle measurements Atmospheric measurements Visualization

来源：评论

学校读者我要写书评

暂无评论

3D Gaze Estimation Based on Facial Feature Tracking

3D Gaze Estimation Based on Facial Feature Tracking

引用

2011 4th IEEE International Conference on Computer Science and Information Technology(ICCSIT 2011)

作者： Man Yi Zhao Xinbo Zhang Ke Shaanxi Provincial Key Lab.of Speech & Image Information Processing Northwestern Polytechnical University

ISBN: (纸本)9781612848334

A 3D gaze estimation and tracking algorithm based on facial feature tracking is presented in this ***,we used the Active Shape Model(ASM) to extract facial feature points with a stereo ***,the full 3D pose of head is estimated by comparing the feature points of current pose with initial head *** that,the center of eyeball is obtained based on a 3D eye model which is related to head pose and the middle point of eye ***,optical axis was computed as 3D vectors through the center of eyeball to the center of ***,visual axis was gotten by adding an angle to optical ***,in our system,a one-time personal calibration is used to determine this *** experimental results show the accuracy of our gaze tracking system achieves less than 3 degree.

关键词： gaze tracking gaze estimation head pose facial feature tracking stereo cameras

来源：评论

学校读者我要写书评

暂无评论

An Investigation on Mutual Information for the Linear Predictive System and the Extrapolation of Speech Signals

An Investigation on Mutual Information for the Linear Predic...

引用

Speech communication;10. ITG Symposium

作者： Jalal Taghia Rainer Martin Jalil Taghia Arne Leijon Institute of Communication Acoustics Ruhr-Universitat Bochum Sound and Image Processing Lab KTH Royal Institute of Technology

ISBN: (纸本)9783800734559

Mutual information (MI) is an important information theoretic concept which has many applications in telecommunications, in blind source separation, and in machine learning. More recently, it has been also employed for the instrumental assessment of speech intelligibility where traditionally correlation based measures are used. In this paper, we address the difference between MI and correlation from the viewpoint of discovering dependencies between variables in the context of speech signals. We perform our investigation by considering the linear predictive approximation and the extrapolation of speech signals as examples. We compare a parametric MI estimation approach based on a Gaussian mixture model (GMM) with the knearest neighbor (KNN) approach which is a well-known non-parametric method available to estimate the MI. We show that the GMM-based MI estimator leads to more consistent results.

关键词： Speech Correlation Extrapolation Mutual information Estimation Speech processing Probability distribution

来源：评论

学校读者我要写书评

暂无评论

Modelling of random extra pulses during quasi-closed glottal cycle phases 10

Modelling of random extra pulses during quasi-closed glottal...

引用

10th International Workshop on Models and Analysis of Vocal Emissions for Biomedical Applications, MAVEBA 2017

作者： Aichinger, P. Roesner, I. Schoentgen, J. Pernkopf, F. Department of Otorhinolaryngology Division of Phoniatrics-Logopedics Medical University of Vienna Austria F.N.R.S. Université Libre de Bruxelles Laboratories of Image Signal Processing and Acoustics Faculty of Applied Sciences Brussels Belgium Signal Processing and Speech Communication Lab Graz University of Technology Austria

ISBN: (纸本)9788864536064

The presence of random extra pulses during quasi-closed glottal cycle phases may constitute a distinct voice quality type relevant to the clinical care of disordered voices. In this paper, we propose for this voice type a glottal area waveform model that includes automatic parameter estimation. The model involves (1) extraction of the fundamental frequency, (2) estimation of the cyclic pulse times, heights and shapes, (3) Fourier synthesis of a cyclic pulse train model, (4) closed phases estimation via fitting an inverted parabola to the averaged pulse shape, (5) estimation of the random extra pulses’ positions and shapes, and (6) pulse shape filtering based synthesis of the random extra pulses. For a typical voice sample, the root mean square error energy level of the purely cyclic model = -13.2 dB, which improves by 1.5 dB when extra pulses are added to the model. © Models and Analysis of Vocal Emissions for Biomedical Applications, MAVEBA *** right reserved.

关键词： Mean square error

来源：评论

学校读者我要写书评

暂无评论

Top-down saliency by multi-scale contextual pooling

Top-down saliency by multi-scale contextual pooling

引用

13th Pacific-Rim Conference on Multimedia, PCM 2012

作者： Qiu, Yuanyuan Zhu, Jun Zhang, Rui Huang, Jun Institute of Image Communication and Network Engineering Shanghai Jiao Tong University Shanghai China Shanghai Key Laboratory of Digital Media Processing and Transmission Shanghai Jiao Tong University Shanghai China Shanghai Advanced Research Institute Chinese Academy of Sciences China

ISBN: (纸本)9783642347771

Goal-driven top-down mechanism plays an important role in the case of object detection and recognition. In this paper, we propose a top-down computational model for goal-driven saliency detection based on a coding-based classification framework. It consists of four successive steps: feature extraction, descriptor coding, local pooling and saliency prediction. In the step of local pooling, we investigate the effect of multi-scale contextual information for saliency detection and find that there exists an optimal contextual scale to achieve the patch-level feature presentation. On basis of this observation, we propose an approach for automatic scale selection in saliency prediction step. The experimental results demonstrate that our method can effectively improve the performance of goal-driven saliency detection as well as related object detection. © 2012 Springer-Verlag.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

Identifying and pruning redundant structures for deep neural networks 34

Identifying and pruning redundant structures for deep neural...

引用

34th IEEE International Conference on Visual communications and image processing, VCIP 2019

作者： Gan, Wenyao Song, Li Chen, Li Xie, Rong Gu, Xiao Shanghai Jiao Tong University Institute of Image Communication and Network Engineering China Shanghai Institute for Advanced Communication and Data Science Shanghai200240 China MoE Key Lab of Artificial Intelligence AI Institute Shanghai Jiao Tong University Shanghai China

ISBN: (纸本)9781728137230

Deep convolutional neural networks have achieved considerable success in the field of computer vision. However, it is difficult to deploy state-of-The-Art models on resource-constrained platforms due to their high storage, memory bandwidth, and computational costs. In this paper, we propose a structured pruning method which employs a three-step process to reduce the resource consumption of neural networks. First, we train an initial network on the training set and evaluate it on the validation set. Next, we introduce an iterative pruning and fine-Tuning algorithm to identify and prune redundant structures, which results in a pruned network with a compact architecture. Finally, we train the pruned network from scratch on both the training set and validation set to obtain the final accuracy on the test set. In the experiments, our pruning method significantly reduces the model size (by 87.2% on CIFAR-10), saves inference time (53.3% on CIFAR-10), and achieves better performance as compared to recent state-of-The-Art methods. © 2019 IEEE.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Co-cyclic jacket matrix based optimal training design and placement for MIMO OFDM channel estimation

Co-cyclic jacket matrix based optimal training design and pl...

引用

2008 4th IEEE International Conference on Circuits and Systems for communications, ICCSC

作者： Lin, Chengyu Zhang, Wenjun Guo, Ying Xu, Youyun Institute of Image Communication and Information Processing Department of Electronic Engineering Shanghai Jiao Tong University Shanghai 200240 China Shanghai Key Laboratory of Digital Media Processing and Transmissions Shanghai Jiao Tong University Shanghai 200240 China School of Information Science and Engineering Central South University Changsha 410083 China

ISBN: (纸本)9781424417087

This paper deals with optimal training design and placement over multiple OFDM symbols for the least squares (LS) channel estimation in multiple-input multiple-output (MIMO) orthogonal frequency division multiplexing (OFDM) systems. When the number of transmit antennas increases, more pilot tones are required and the transmission efficiency decreases. To solve this problem, the pilot sequences are distributed over multiple OFDM symbols. Motivated by the properties of co-cyclic Jacket matrices, we derive the optimal pilot sequences from co-cyclic Jacket matrices based on the minimum mean square error (MSE) of the LS channel estimation. Simulation results show that the derived optimal pilot sequences over one, two and four OFDM symbols have the same performance, and greatly outperform the orthogonal pilot sequences in slowly time-varying channels. ©2008 IEEE.

关键词： Orthogonal frequency division multiplexing

来源：评论

学校读者我要写书评

暂无评论

Physics-Environment Interaction Network for Dense Crowd Behavior Recognition

SSRN

引用

SSRN 2024年

作者： Yu, Jiaqi Zhou, Yanshan Pan, Renjie Lai, Pingrui Yang, Hua Institute of Image Communication and Network Engineering Shanghai Key Lab of Digital Media Processing and Transmission Shanghai Jiao Tong University Shanghai200240 China

The analysis of large-scale crowd behavior plays a crucial role in public safety. However, intelligent systems face three major challenges in analyzing dense crowd behavior: the severe occlusion between individuals, the variability in behavior patterns, and the complexity of behavioral evolution. To address these challenges, we propose the Physics-Environment Interaction Network (PEIN), which directly models the motion characteristics of a group with its physics attributes. Specifically, our method consists of two streams. The first stream is the physics-informed crowd property stream, which leverages on the similarity between dense crowd motion and fluid dynamics, using the Navier-Stokes (N-S) equation from fluid mechanics as the modeling framework to describe crowd motion. Considering the inherent relationship between the terms in the N-S equation and various crowd properties (collectiveness, conflict, uniformity and stability), we model these terms with operators and neural networks guided by these crowd properties, enabling the modeling of crowd motion characteristic without relying on the extraction of individual motion information. The second stream is the environment perception stream. Considering that the physics-informed crowd property stream mainly focuses on instantaneous information and that scenes with dense crowd behavior have variability, we introduce a 3D network to enhance the model's robustness. This stream can extract global spatiotemporal information from input video frames, providing a comprehensive perception of the surrounding crowd environment. Since the two streams process different data sources, we design a dual cross-attention mechanism to to enable between features from different modalities, resulting in a joint learnable representation for the final crowd behavior recognition. By incorporating physical laws as constraints, we design a physics-informed loss function combined with a crowd behavior loss function to optimize the model. Consi

关键词： Navier Stokes equations

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：