检索结果-内蒙古大学图书馆

Self-grouping convolutional neural networks

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Guo, Qingbei Wu, Xiao-Jun Kittler, Josef Feng, Zhiquan Jiangsu Provincial Engineering Laboratory of Pattern Recognition and Computational Intelligence Jiangnan University Wuxi214122 China Shandong Provincial Key Laboratory of Network based Intelligent Computing University of Jinan Jinan250022 China Centre for Vision Speech and Signal Processing University of Surrey GuildfordGU2 7XH United Kingdom

Although group convolution operators are increasingly used in deep convolutional neural networks to improve the computational efficiency and to reduce the number of parameters, most existing methods construct their group convolution architectures by a predefined partitioning of the filters of each convolutional layer into multiple regular filter groups with an equal spatial group size and data-independence, which prevents a full exploitation of their potential. To tackle this issue, we propose a novel method of designing self-grouping convolutional neural networks, called SG-CNN, in which the filters of each convolutional layer group themselves based on the similarity of their importance vectors. Concretely, for each filter, we first evaluate the importance value of their input channels to identify the importance vectors, and then group these vectors by clustering. Using the resulting data-dependent centroids, we prune the less important connections, which implicitly minimizes the accuracy loss of the pruning, thus yielding a set of diverse group convolution filters. Subsequently, we develop two fine-tuning schemes, i.e. (1) both local and global fine-tuning and (2) global only fine-tuning, which experimentally deliver comparable results, to recover the recognition capacity of the pruned network. Comprehensive experiments carried out on the CIFAR-10/100 and ImageNet datasets demonstrate that our self-grouping convolution method adapts to various state-of-the-art CNN architectures, (Figure presented) Figure 1: Evolution of group convolutions. (a) Regular convolution. (b) Regular group convolution. (c) Permuting group convolution. (d) Learned group convolution. (d) Self-grouping convolution. Note that white channels represent the ignored input channels, and gray channels indicate the reused input channels. such as ResNet and DenseNet, and delivers superior performance in terms of compression ratio, speedup and recognition accuracy. We demonstrate the ability of SG-CNN

关键词： Convolution

Active Lighting Recurrence by Parallel Lighting Analogy for Fine-Grained Change Detection

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Zhang, Qian Feng, Wei Wan, Liang Tian, Fei-Peng Wang, Xiaowei Tan, Ping School of Computer Science and Technology College of Intelligence and Computing Tianjin University TianJin300350 China State Administration of Cultural Heritage China School of Computer Software College of Intelligence and Computing Tianjin University TianJin300350 China Tianjin Key Lab for Advanced Signal Processing Civil Aviation University of China Dunhuang Research Academy Dunhuang Gansu China School of Computing Science Simon Fraser University Canada

—This paper studies a new problem, namely active lighting recurrence (ALR) that physically relocalizes a light source to reproduce the lighting condition from single reference image for a same scene, which may suffer from fine-grained changes during twice observations. ALR is of great importance for fine-grained visual inspection and change detection, because some phenomena or minute changes can only be clearly observed under particular lighting conditions. Therefore, effective ALR should be able to online navigate a light source toward the target pose, which is challenging due to the complexity and diversity of real-world lighting and imaging processes. To this end, we propose to use the simple parallel lighting as an analogy model and based on Lambertian law to compose an instant navigation ball for this purpose. We theoretically prove the feasibility, i.e., equivalence and convergence, of this ALR approach for realistic near point light source and small near surface light source. Besides, we also theoretically prove the invariance of our ALR approach to the ambiguity of normal and lighting decomposition. The effectiveness and superiority of the proposed approach have been verified by both extensive quantitative experiments and challenging real-world tasks on fine-grained change detection of cultural heritages. We also validate the generality of our approach to non-Lambertian scenes. Copyright © 2020, The Authors. All rights reserved.

关键词： Light sources

Hard instance learning for quantum adiabatic prime factorization

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Lin, Jian Zhang, Zhengfeng Zhang, Junping Li, Xiaopeng State Key Laboratory of Surface Physics Institute of Nanoelectronics and Quantum Computing Department of Physics Fudan University Shanghai200433 China Shanghai Key Lab of Intelligent Information Processing School of Computer Science Fudan University Shanghai200433 China Shanghai Qi Zhi Institute AI Tower Xuhui District Shanghai200232 China

Prime factorization is a difficult problem with classical computing, whose exponential hardness is the foundation of Rivest-Shamir-Adleman (RSA) cryptography. With programmable quantum devices, adiabatic quantum computing has been proposed as a plausible approach to solve prime factorization, having promising advantage over classical computing. Here, we find there are certain hard instances that are consistently intractable for both classical simulated annealing and un-configured adiabatic quantum computing (AQC). Aiming at an automated architecture for optimal configuration of quantum adiabatic factorization, we apply a deep reinforcement learning (RL) method to configure the AQC algorithm. By setting the success probability of the worst-case problem instances as the reward to RL, we show the AQC performance on the hard instances is dramatically improved by RL configuration. The success probability also becomes more evenly distributed over different problem instances, meaning the configured AQC is more stable as compared to the un-configured case. Through a technique of transfer learning, we find prominent evidence that the framework of AQC configuration is scalable—the configured AQC as trained on five qubits remains working efficiently on nine qubits with a minimal amount of additional training cost. Copyright © 2021, The Authors. All rights reserved.

关键词： Factorization

Robust DOA Estimation Based on Dual Lawson Norm for RIS-Aided Wireless Communication Systems

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Yu, Canping Li, Yingsong Li, Liping Huang, Zhixiang Wu, Qingqing de Lamare, Rodrigo C. The Key Laboratory of Intelligent Computing and Signal Processing Ministry of Education Anhui University Anhui Hefei230601 China The Department of Electronic Engineering Shanghai Jiao Tong University 200240 China Gávea 22451-900 Brazil The School of Physics Engineering and Technology University of York YorkYO10 5DD United Kingdom

Reconfigurable intelligent surfaces (RIS) can actively perform beamforming and have become a crucial enabler for wireless systems in the future. The direction-of-arrival (DOA) estimates of RIS received signals can help design the reflection control matrix and improve communication quality. In this paper, we design a RIS-assisted system and propose a robust Lawson norm-based multiple-signal-classification (LN-MUSIC) DOA estimation algorithm for impulsive noise environments, which is divided into two parts: first, the non-convex Lawson norm is used as the error criterion along with a regularization constraint to formulate the optimization problem. Then, a Bregman distance based alternating-direction-method-of-multipliers (ADMM) is used to solve the problem and recover the desired signal. The second part is to use the multiple-signal-classification (MUSIC) to find out the DOAs of targets based on their sparsity in the spatial domain. In addition, we also propose a RIS control matrix optimization strategy that requires no channel state information (CSI), which effectively enhances the desired signals and improves the performance of the LN-MUSIC algorithm. A Cramér–Rao-lower-bound (CRLB) of the proposed DOA estimation algorithm is presented and verifies its feasibility. Simulated results show that the proposed robust DOA estimate algorithm based on the Lawson norm can effectively suppress the impact of large outliers caused by impulsive noise on the estimation results, outperforming existing *** Codes - Copyright © 2024, The Authors. All rights reserved.

关键词： Impulse noise

The Application of the ADI-FDTD method in the 3D hybrid subgrid technique

学校读者我要写书评

暂无评论

The Application of the ADI-FDTD method in the 3D hybrid subg...

International Applied Computational Electromagnetics Society Symposium (ACES)

作者： Ziheng Song Guoda Xie Zhixiang Huang Guilin Hou Pan Pan School of Electronic and Information Engineering Anhui university Hefei China Education of Anhui Province. Anhui Province Key Laboratory of Target Recognition and Feature Extraction (Luan Anhui 237000 China) Ministry of Education. the Key Laboratory of Intelligent Computing and Signal Processing Hefei China Anhui Province Anhui University. Key Laboratory of Electromagnetic Environmental Sensing (Hefei 230601 China) Anhui university Hefei China

A hybrid subgrid scheme based on the conventional finite-difference time-domain (FDTD) schemes are proposed. The alternating-direction-implicit FDTD (ADI-FDTD) is used to calculate electromagnetic fields in fine grid regions subgrids and the FDTD method is applied to the coarse grid regions. Numerical results demonstrate that the drastically reduced computation time and memory can be achieved by using the proposed 3D hybrid algorithm.

关键词： Three-dimensional displays Costs Computational modeling Computational electromagnetics Electromagnetic fields Time-domain analysis Finite difference methods

A Theoretically Guaranteed Quaternion Weighted Schatten p-norm Minimization Method for Color Image Restoration

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Zhang, Qing-Hua He, Liang-Tian Wang, Yi-Lun Deng, Liang-Jian Liu, Jun Key Laboratory of Intelligent Computing and Signal Processing of Ministry of Education School of Mathematical Sciences Anhui University Hefei230601 China Machinify Inc Palo AltoCA94301 United States School of Mathematical Sciences University of Electronic Science and Technology of China Chengdu611731 China Key Laboratory for Applied Statistics of MOE School of Mathematics and Statistics Northeast Normal University Changchun130024 China

Inspired by the fact that the matrix formulated by nonlocal similar patches in a natural image is of low rank, the rank approximation issue have been extensively investigated over the past decades, among which weighted nuclear norm minimization (WNNM) and weighted Schatten p-norm minimization (WSNM) are two prevailing methods have shown great superiority in various image restoration (IR) problems. Due to the physical characteristic of color images, color image restoration (CIR) is often a much more difficult task than its grayscale image counterpart. However, when applied to CIR, the traditional WNNM/WSNM method only processes three color channels individually and fails to consider their cross-channel correlations. Very recently, a quaternion-based WNNM approach (QWNNM) has been developed to mitigate this issue, which is capable of representing the color image as a whole in the quaternion domain and preserving the inherent correlation among the three color channels. Despite its empirical success, unfortunately, the convergence behavior of QWNNM has not been strictly studied yet. In this paper, on the one side, we extend the WSNM into quaternion domain and correspondingly propose a novel quaternion-based WSNM model (QWSNM) for tackling the CIR problems. Extensive experiments on two representative CIR tasks, including color image denoising and deblurring, demonstrate that the proposed QWSNM method performs favorably against many state-of-the-art alternatives, in both quantitative and qualitative evaluations. On the other side, more importantly, we preliminarily provide a theoretical convergence analysis, that is, by modifying the quaternion alternating direction method of multipliers (QADMM) through a simple continuation strategy, we theoretically prove that both the solution sequences generated by the QWNNM and QWSNM have fixed-point convergence guarantees. The source code of our algorithm can be downloaded at the website: https://***/qiuxuanzhizi/QWSNM. Copyr

关键词： Color

3D LOD-FDTD for the Multi-Pole Debye Dispersive Media Using Z-Transform

学校读者我要写书评

暂无评论

3D LOD-FDTD for the Multi-Pole Debye Dispersive Media Using ...

International Applied Computational Electromagnetics Society Symposium (ACES)

作者： Guilin Hou Ziheng Song Guoda Xie Pan Pan Zhixiang Huang School of Electronic and Information Engineering Anhui university Hefei China Education of Anhui Province. Anhui Province Key Laboratory of Target Recognition and Feature Extraction (Luan Anhui 237000 China) Ministry of Education. the Key Laboratory of Intelligent Computing and Signal Processing Hefei China Anhui Province Anhui University. Key Laboratory of Electromagnetic Environmental Sensing (Hefei 230601 China) Anhui university Hefei China

This paper introduces the application of unconditionally stable locally one-dimensional finite-difference time-domain(LOD-FDTD) method to 3D multi-pole Debye dispersive media model. Compared with other methods for dispersive media, the Z-transform method does not need to retain the electromagnetic values of the previous moments, has less memory footprint, and its numerical implementation is more straightforward. Through numerical examples, ZT-LOD-FDTD shows the advantages of accuracy and high computational efficiency.

关键词： Solid modeling Three-dimensional displays Memory management Media Electromagnetic modeling Numerical models Time-domain analysis

Several constructions of optimal LCD codes over small finite fields

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Li, Shitao Shi, Minjia Liu, Huizhou School of Mathematical Sciences Anhui University Hefei230601 China Key Laboratory of Intelligent Computing and Signal Processing Ministry of Education School of Mathematical Sciences Anhui University Hefei230601 China State Key Laboratory of Information Security Institute of Information Engineering Chinese Academy of Sciences Beijing100093 China State Grid Anhui Electric Power Co. Ltd. Hefei China

Linear complementary dual (LCD) codes are linear codes which intersect their dual codes trivially, which have been of interest and extensively studied due to their practical applications in computational complexity and information protection. In this paper, we give some methods for constructing LCD codes over small finite fields by modifying some typical methods for constructing linear codes. We show that all odd-like binary LCD codes, ternary LCD codes and quaternary Hermitian LCD codes can be constructed using the modified methods. Our results improve the known lower bounds on the largest minimum distances of LCD codes. Furthermore, we give two counterexamples to disprove the conjecture proposed by Bouyuklieva (Des. Codes Cryptogr. 89(11): 2445-2461, 2021). Copyright © 2022, The Authors. All rights reserved.

关键词：

Method of ICA based on estimating the PDF of signals

学校读者我要写书评

暂无评论

Xi'an Dianzi Keji Daxue Xuebao/Journal of Xidian University 2005年第4期32卷 574-578页

作者： Li, Xiao-Jun Lou, Shun-Tian Zhang, Xian-Da Key Lab. of Radar Signal Processing Xidian Univ. Xi'an 710071 China State Key Lab. of Intelligent Technology and Systems Dept. of Automation Tsinghua Univ. Beijing 100084 China

Independent component analysis (ICA) is a method for finding independent components from multivariate (multidimensional) statistical data. Based on the optimal estimation function, a method for the estimation of the score function is developed. By using the Gaussian mixture model, an EM algorithm for approximating the probability density of the data is presented, and a stochastic gradient method is given to separate the independent components. To improve the accuracy and stability of the algorithm, an iterative method for estimating the PDF of data is presented, which can perform the separation of mixed sub-Gaussian from super-Gaussian sources. The performance of the method is shown by computer simulations.

关键词： Independent component analysis