检索结果-南通市图书馆

Orbital angular momentum-mediated machine learning for high-accuracy mode-feature encoding

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Light(science & Applications) 2024年第3期13卷 466-477页

作者： Xinyuan Fang Xiaonan Hu Baoli Li Hang Su Ke Cheng Haitao Luan Min Gu institute of Photonic Chips University of Shanghai for Science and TechnologyShanghai 200093China Centre for artificial-intelligence Nanophotonics School of Optical-Electrical and Computer EngineeringUniversity of Shanghai for Science and TechnologyShanghai 200093China

Machine learning with optical neural networks has featured unique advantages of the information processing including high speed,ultrawide bandwidths and low energy consumption because the optical dimensions(time,space,wavelength,and polarization)could be utilized to increase the degree of freedom.However,due to the lack of the capability to extract the information features in the orbital angular momentum(OAM)domain,the theoretically unlimited OAM states have never been exploited to represent the signal of the input/output nodes in the neural network model.Here,we demonstrate OAM-mediated machine learning with an all-optical convolutional neural network(CNN)based on Laguerre-Gaussian(LG)beam modes with diverse diffraction losses.The proposed CNN architecture is composed of a trainable OAM mode-dispersion impulse as a convolutional kernel for feature extraction,and deep-learning diffractive layers as a classifier.The resultant OAM mode-dispersion selectivity can be applied in information mode-feature encoding,leading to an accuracy as high as 97.2%for MNIST database through detecting the energy weighting coefficients of the encoded OAM modes,as well as a resistance to eavesdropping in point-to-point free-space transmission.Moreover,through extending the target encoded modes into multiplexed OAM states,we realize all-optical dimension reduction for anomaly detection with an accuracy of 85%.Our work provides a deep insight to the mechanism of machine learning with spatial modes basis,which can be further utilized to improve the performances of various machine-vision tasks by constructing the unsupervised learning-based auto-encoder.

关键词： polarization mode momentum

Joint Sample Position Based Noise Filtering and Mean Shift Clustering for Imbalanced Classification Learning

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Tsinghua science and technology 2024年第1期29卷 216-231页

作者： Lilong Duan Wei Xue Jun Huang Xiao Zheng School of computer science and technology Anhui University of TechnologyMaanshan 243032China institute of artificial intelligence Hefei Comprehensive National Science CenterHefei 230088China

The problem of imbalanced data classification learning has received much attention.Conventional classification algorithms are susceptible to data skew to favor majority samples and ignore minority samples.Majority weighted minority oversampling technique(MWMOTE)is an effective approach to solve this problem,however,it may suffer from the shortcomings of inadequate noise filtering and synthesizing the same samples as the original minority data.To this end,we propose an improved MWMOTE method named joint sample position based noise filtering and mean shift clustering(SPMSC)to solve these problems.Firstly,in order to effectively eliminate the effect of noisy samples,SPMsC uses a new noise filtering mechanism to determine whether a minority sample is noisy or not based on its position and distribution relative to the majority sample.Note that MWMOTE may generate duplicate samples,we then employ the mean shift algorithm to cluster minority samples to reduce synthetic replicate samples.Finally,data cleaning is performed on the processed data to further eliminate class overlap.Experiments on extensive benchmark datasets demonstrate the effectiveness of SPMsC compared with other sampling methods.

关键词： imbalanced data classification oversampling noise filtering clustering

Understanding adversarial attacks on observations in deep reinforcement learning

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

science China(Information sciences) 2024年第5期67卷 69-83页

作者： You QIAOBEN Chengyang YING Xinning ZHOU Hang SU Jun ZHU Bo ZHANG department of computer science and technology Beijing National Research Center for Information Science and Technology Tsinghua-Bosch Joint Center for Machine Learning Institute for Artificial Intelligence Tsinghua University Peng Cheng Laboratory

Deep reinforcement learning models are vulnerable to adversarial attacks that can decrease the cumulative expected reward of a victim by manipulating its observations. Despite the efficiency of previous optimization-based methods for generating adversarial noise in supervised learning, such methods might not achieve the lowest cumulative reward since they do not generally explore the environmental dynamics.Herein, a framework is provided to better understand the existing methods by reformulating the problem of adversarial attacks on reinforcement learning in the function space. The reformulation approach adopted herein generates an optimal adversary in the function space of targeted attacks, repelling them via a generic two-stage framework. In the first stage, a deceptive policy is trained by hacking the environment and discovering a set of trajectories routing to the lowest reward or the worst-case performance. Next, the adversary misleads the victim to imitate the deceptive policy by perturbing the observations. Compared to existing approaches, it is theoretically shown that our adversary is strong under an appropriate noise level. Extensive experiments demonstrate the superiority of the proposed method in terms of efficiency and effectiveness, achieving state-of-the-art performance in both Atari and MuJoCo environments.

关键词： deep learning reinforcement learning adversarial robustness adversarial attack

Retrieval-and-alignment based large-scale indoor point cloud semantic segmentation

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

science China(Information sciences) 2024年第4期67卷 164-180页

作者： Zongyi XU Xiaoshui HUANG Bo YUAN Yangfu WANG Qianni ZHANG Weisheng LI Xinbo GAO School of computer science and technology Chongqing University of Posts and Telecommunications Chongqing institute for Brain and intelligence Guangyang Bay Laboratory 3. Shanghai Artificial Intelligence Laboratory School of Electronic Engineering and computer science Queen Mary University of London

Current methods for point cloud semantic segmentation depend on the extraction of descriptive features. However, unlike images, point clouds are irregular and often lack texture information, making it demanding to extract discriminative features. In addition, noise, outliers, and uneven point distribution are commonly present in point clouds, which further complicates the segmentation task. To address these problems, a novel architecture is proposed for direct and accurate large-scale point cloud segmentation based on point cloud retrieval and alignment. The proposed approach involves using a feature-based point cloud retrieval method for searching for reference point clouds with annotations from a dataset. In the following segmentation stage, an overlap-based point cloud registration method has been developed to align the target and reference point clouds. For accurate and robust alignment, an overlap region estimation module is trained to locate the optimal overlap region between two pieces of point clouds in a coarse-to-fine manner. In the detected overlap region, the global and local features of the points are extracted and combined for featuremetric registration to obtain accurate transformation parameters between the target and reference point clouds. After alignment, the annotated segmentation of the reference is transferred to the target point clouds to obtain accurate segmentation results. Extensive experiments are conducted to show that the developed method outperforms the state-of-the-art approaches in terms of both accuracy and robustness against noise and outliers.

关键词： point cloud semantic segmentation large-scale indoor point clouds point cloud alignment overlap estimation label transfer

Robust Regularization Design of Graph Neural Networks Against Adversarial Attacks Based on Lyapunov Theory

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Chinese Journal of Electronics 2024年第3期33卷 732-741页

作者： Wenjie YAN Ziqi LI Yongjun QI School of artificial intelligence Hebei University of Technology School of computer science and Engineering of North China institute of Aerospace Engineering

The robustness of graph neural networks（GNNs） is a critical research topic in deep learning.Many researchers have designed regularization methods to enhance the robustness of neural networks,but there is a lack of theoretical analysis on the principle of robustness.In order to tackle the weakness of current robustness designing methods,this paper gives new insights into how to guarantee the robustness of GNNs.A novel regularization strategy named Lya-Reg is designed to guarantee the robustness of GNNs by Lyapunov theory.Our results give new insights into how regularization can mitigate the various adversarial effects on different graph signals.Extensive experiments on various public datasets demonstrate that the proposed regularization method is more robust than the state-of-theart methods such as L1-norm,L2-norm,L2-norm,Pro-GNN,PA-GNN and GARNET against various types of graph adversarial attacks.

关键词： Training Deep learning Garnets Robustness Graph neural networks

Ultrahigh performance passive radiative cooling by hybrid polar dielectric metasurface thermal emitters

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Opto-Electronic Advances 2024年第4期7卷 17-25页

作者： Yinan Zhang Yinggang Chen Tong Wang Qian Zhu Min Gu institute of Photonic Chips University of Shanghai for Science and TechnologyShanghai 200093China Centre for artificial-intelligence Nanophotonics School of Optical-Electrical and Computer EngineeringUniversity of Shanghai for Science and TechnologyShanghai 200093China

Real-world passive radiative cooling requires highly emissive,selective,and omnidirectional thermal emitters to maintain the radiative cooler at a certain temperature below the ambient temperature while maximizing the net cooling power.Despite various selective thermal emitters have been demonstrated,it is still challenging to achieve these conditions sim-ultaneously because of the extreme difficulty in controlling thermal emission of photonic structures in multidimension.Here we demonstrated hybrid polar dielectric metasurface thermal emitters with machine learning inverse design,en-abling a high emissivity of~0.92 within the atmospheric transparency window 8-13μm,a large spectral selectivity of~1.8 and a wide emission angle up to 80 degrees,simultaneously.This selective and omnidirectional thermal emitter has led to a new record of temperature reduction as large as~15.4°C under strong solar irradiation of~800 W/m2,signific-antly surpassing the state-of-the-art results.The designed structures also show great potential in tackling the urban heat island effect,with modelling results suggesting a large energy saving and deployment area reduction.This research will make significant impact on passive radiative cooling,thermal energy photonics and tackling global climate change.

关键词： radiative cooling dielectric metasurfaces machine learning thermal emitters

维普期刊数据库

Image enhancement with intensity transformation on embedding space

在线全文

学校读者我要写书评

暂无评论

CAAI Transactions on intelligence technology 2024年第1期9卷 101-115页

作者： Hanul Kim Yeji Jeon Yeong Jun Koh department of Applied artificial intelligence Seoul National University of Science and TechnologySeoulSouth Korea department of computer science and Engineering Chungnam National UniversityDaejeonSouth Korea

In recent times,an image enhancement approach,which learns the global transformation function using deep neural networks,has gained attention.However,many existing methods based on this approach have a limitation:their transformation functions are too simple to imitate complex colour transformations between low-quality images and manually retouched high-quality images.In order to address this limitation,a simple yet effective approach for image enhancement is proposed.The proposed algorithm based on the channel-wise intensity transformation is designed.However,this transformation is applied to the learnt embedding space instead of specific colour spaces and then return enhanced features to colours.To this end,the authors define the continuous intensity transformation(CIT)to describe the mapping between input and output intensities on the embedding space.Then,the enhancement network is developed,which produces multi-scale feature maps from input images,derives the set of transformation functions,and performs the CIT to obtain enhanced images.Extensive experiments on the MIT-Adobe 5K dataset demonstrate that the authors’approach improves the performance of conventional intensity transforms on colour space metrics.Specifically,the authors achieved a 3.8%improvement in peak signal-to-noise ratio,a 1.8%improvement in structual similarity index measure,and a 27.5%improvement in learned perceptual image patch similarity.Also,the authors’algorithm outperforms state-of-the-art alternatives on three image enhancement datasets:MIT-Adobe 5K,Low-Light,and Google HDRþ.

关键词： computer vision deep learning image enhancement image processing

A Privacy Preservation Method for Attributed Social Network Based on Negative Representation of Information

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

computer Modeling in Engineering & sciences 2024年第7期140卷 1045-1075页

作者： Hao Jiang Yuerong Liao Dongdong Zhao Wenjian Luo Xingyi Zhang Key Laboratory of Intelligent Computing and Signal Processing of Ministry of Education School of Computer Science and TechnologyAnhui UniversityHefei230601China Chongqing Research institute School of Computer Science and Artificial IntelligenceWuhan University of TechnologyWuhan430070China Guangdong Provincial Key Laboratory of Novel Security intelligence Technologies School of Computer Science and TechnologyHarbin Institute of TechnologyShenzhen518055China

Due to the presence of a large amount of personal sensitive information in social networks,privacy preservation issues in social networks have attracted the attention of many scholars.Inspired by the self-nonself discrimination paradigmin the biological immune system,the negative representation of information indicates features such as simplicity and efficiency,which is very suitable for preserving social network privacy.Therefore,we suggest a method to preserve the topology privacy and node attribute privacy of attribute social networks,called AttNetNRI.Specifically,a negative survey-based method is developed to disturb the relationship between nodes in the social network so that the topology structure can be kept private.Moreover,a negative database-based method is proposed to hide node attributes,so that the privacy of node attributes can be preserved while supporting the similarity estimation between different node attributes,which is crucial to the analysis of social networks.To evaluate the performance of the AttNetNRI,empirical studies have been conducted on various attribute social networks and compared with several state-of-the-art methods tailored to preserve the privacy of social networks.The experimental results show the superiority of the developed method in preserving the privacy of attribute social networks and demonstrate the effectiveness of the topology disturbing and attribute hiding parts.The experimental results show the superiority of the developed methods in preserving the privacy of attribute social networks and demonstrate the effectiveness of the topological interference and attribute-hiding components.

关键词： Attributed social network topology privacy node attribute privacy negative representation of information negative survey negative database

Sparse representation scheme with enhanced medium pixel intensity for face recognition

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

CAAI Transactions on intelligence technology 2024年第1期9卷 116-127页

作者： Xuexue Zhang Yongjun Zhang Zewei Wang Wei Long Weihao Gao Bob Zhang State Key Laboratory of Public Big Data Institute for Artificial IntelligenceCollege of Computer Science and TechnologyGuizhou UniversityGuiyangGuizhouChina department of computer and Information science University of MacaoAvenida da UniversidadeTaipaMacaoChina

Sparse representation is an effective data classification algorithm that depends on the known training samples to categorise the test sample.It has been widely used in various image classification tasks.Sparseness in sparse representation means that only a few of instances selected from all training samples can effectively convey the essential class-specific information of the test sample,which is very important for classification.For deformable images such as human faces,pixels at the same location of different images of the same subject usually have different intensities.Therefore,extracting features and correctly classifying such deformable objects is very hard.Moreover,the lighting,attitude and occlusion cause more difficulty.Considering the problems and challenges listed above,a novel image representation and classification algorithm is proposed.First,the authors’algorithm generates virtual samples by a non-linear variation method.This method can effectively extract the low-frequency information of space-domain features of the original image,which is very useful for representing deformable objects.The combination of the original and virtual samples is more beneficial to improve the clas-sification performance and robustness of the algorithm.Thereby,the authors’algorithm calculates the expression coefficients of the original and virtual samples separately using the sparse representation principle and obtains the final score by a designed efficient score fusion scheme.The weighting coefficients in the score fusion scheme are set entirely automatically.Finally,the algorithm classifies the samples based on the final scores.The experimental results show that our method performs better classification than conventional sparse representation algorithms.

关键词： computer vision face recognition image classification image representation