检索结果-南通市图书馆

convolutional neural networks with Large-Margin Softmax Loss Function for Cognitive Load Recognition 36

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Convolutional Neural Networks with Large-Margin Softmax Loss...

Connected Components-based Colour Image Representations of Vibrations for a Two-stage Fault Diagnosis of Roller Bearings Using convolutional neural networks

第36届中国控制会议

作者： Yuetian Liu Qingshan Liu School of Automation Huazhong University of Science and Technology

ISBN: (纸本)9781538629185

Cognitive load recognition has been widely studied recently,but how to find the effective and robust feature representations from the electroencephalography（EEG） signals is still a *** this paper we design lightweight 1D and2 D convolutional neural networks（CNNs） with large-margin softmax loss functions for cognitive load ***,we extract the frequency domain features from the EEG ***,the extracted frequency feature vectors and matrices are used to train the 1D and 2D *** approach approximately achieves 93%accuracy for 1D CNNs and 91%for 2D CNNs with shorter training time on the 4 classes of cognitive load recognition *** results indicate that the proposed method with few parameters gets good performance and is more efficient than other deep learning *** find that the large-margin softmax loss function works well for EEG signal recognition,and we also find that the 1D CNNs with 3×1 convolution kernel have good classification performance with few parameters and fast training rate to be more suitable for EEG signal classification than 2D CNNs for small EEG ***,2D CNNs are suitable for large EEG dataset because of their fast training rate.

关键词： Electroencephalography(EEG) Frequency Domain Features convolutional neural networks Large-margin Softmax Loss Function

来源： cnki会议评论

在线全文

cnki会议

学校读者我要写书评

暂无评论

Chinese Journal of Mechanical Engineering 2021年第3期34卷 73-93页

作者： Hosameldin O.A.Ahmed Asoke K Nandi Department of Electronic and Electrical Engineering Brunel University LondonUxbridge UB83PHUK School of Mechanical Engineering Xi’an Jiaotong UniversityXi’anChina

Roller bearing failure is one of the most common faults in rotating *** techniques for bearing fault diagnosis based on faults feature extraction have been *** feature extraction from fault signals requires expert prior information and human ***,deep learning algorithms have been applied extensively in the condition monitoring of rotating machines to learn features automatically from the input *** its robust performance in image recognition,the convolutional neural network(CNN)architecture has been widely used to learn automatically discriminative features from vibration images and classify health *** paper proposes and evaluates a two-stage method RGBVI-CNN for roller bearings fault *** first stage in the proposed method is to generate the RGB vibration images(RGBVIs)from the input vibration *** begin this process,first,the 1-D vibration signals were converted to 2-D grayscale vibration *** the conversion was completed,the regions of interest(ROI)were found in the converted 2-D grayscale vibration ***,to produce vibration images with more discriminative characteristics,an algorithm was applied to the 2-D grayscale vibration images to produce connected components-based RGB vibration images(RGBVIs)with sets of colours and texture *** the second stage,with these RGBVIs a CNN-based architecture was employed to learn automatically features from the RGBVIs and to classify bearing health *** cases of fault classification of rolling element bearings are used to validate the proposed *** results of this investigation demonstrate that RGBVI-CNN can generate advantageous health condition features from bearing vibration signals and classify the health conditions under different working loads with high ***,several classification models trained using RGBVI-CNN offered high performance in the testing results of the overall classification accura

关键词： Bearing fault diagnosis Image representation of vibrations Deep learning convolutional neural networks

The Kernel Dynamics of convolutional neural networks in Manifolds

在线全文

学校读者我要写书评

暂无评论

Chinese Journal of Electronics 2020年第6期29卷 1185-1192页

作者： WU Wei JING Xiaoyuan DU Wencai School of Computer Science Wuhan University Institute of Deep-sea Science and Engineering Chinese Academy of Sciences Institute of Data Science City University of Macau

We propose a novel expression from manifolds to define convolutional neural network(CNN).The layered structure is proceeded by integration in limited space continuously,with weights adjusted including value and direction in neural *** transfer functions are proposed to simulate the kernel dynamics as a control *** theoretically analyze the stability and controllability of kernel-based CNNs,and verify our findings by numerical experiments.

关键词： convolutional neural nets learning (artificial intelligence) kernel-based CNN convolutional neural network kernel dynamics status transfer functions neural manifolds layered structure convolutional neural networks Kernelbased convolution Learning dynamics neural manifolds

Exploiting multi-channels deep convolutional neural networks for multivariate time series classification

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Frontiers of Computer Science 2016年第1期10卷 96-112页

作者： Yi ZHENG QiLIU Enhong CHEN Yong GE J. Leon ZHAO School of Computer Science and Technology University of Science and Technology of China Hefei 230027 China Department of Computer Science University of North Carolina at Charlotte Charlotte 28223 USA Department of Information Systems City University of Hong Kong Hong Kong China

Time series classification is related to many dif- ferent domains, such as health informatics, finance, and bioinformatics. Due to its broad applications, researchers have developed many algorithms for this kind of tasks, e.g., multivariate time series classification. Among the classifi- cation algorithms, k-nearest neighbor （k-NN） classification （particularly 1-NN） combined with dynamic time warping （DTW） achieves the state of the art performance. The defi- ciency is that when the data set grows large, the time con- sumption of 1-NN with DTW will be very expensive. In con- trast to 1-NN with DTW, it is more efficient but less ef- fective for feature-based classification methods since their performance usually depends on the quality of hand-crafted features. In this paper, we aim to improve the performance of traditional feature-based approaches through the feature learning techniques. Specifically, we propose a novel deep learning framework, multi-channels deep convolutional neu- ral networks （MC-DCNN）, for multivariate time series classi- fication. This model first learns features from individual uni- variate time series in each channel, and combines information from all channels as feature representation at the final layer. Then, the learnt features are applied into a multilayer percep- tron （MLP） for classification. Finally, the extensive experi- ments on real-world data sets show that our model is not only more efficient than the state of the art but also competitive in accuracy. This study implies that feature learning is worth to be investigated for the problem of time series classification.

关键词： convolutional neural networks time series clas-sification feature learning deep learning

Elastography mapped by deep convolutional neural networks

在线全文

学校读者我要写书评

暂无评论

Science China(Technological Sciences) 2021年第7期64卷 1567-1574页

作者： LIU DongXu KRUGGEL Frithjof SUN LiZhi Department of Civil and Environmental Engineering University of CaliforniaIrvineCA92697USA Department of Biomedical Engineering University of CaliforniaIrvineCA92697USA

Elastography emerges as a medical modality to map stiffness distribution of tissues and is expected to help identify malignant tumors. To this end, tissues are externally stimulated with dynamic waves, and thereafter mechanical responses are internally measured. However, internal measurements limit the resolution and accuracy due to wave scattering and *** models have been reported only with need for acquiring transmitted responses, the computational processes are timeconsuming in the inverse analysis. Here we develop an architecture of deep learning-based convolutional neural networks(CNNs) to image elastography based on sound transmission. The proposed CNNs contain three branches, one of which considers the contribution of original features in input data. By comparison, the developed architecture not only maps elastography accurately, but also is more efficient than traditional CNNs in sequence.

关键词： elastography Young's modulus deep learning convolutional neural networks

Signal-background discrimination with convolutional neural networks in the PandaX-Ⅲ experiment using MC simulation

在线全文

学校读者我要写书评

暂无评论

Science China(Physics,Mechanics & Astronomy) 2018年第10期61卷 51-59页

作者： Hao Qiao ChunYu Lu Xun Chen Ke Han XiangDong Ji SiGuang Wang School of Physics and State Key Laboratory of Nuclear Physics and Technology and Center for High Energy Physics Peking University Beijing 100871 China Institute of Particle and Nuclear Physics and School of Physics and Astronomy Shanghai Jiao Tong University Shanghai Laboratory for Particle Physics and Cosmology Shanghai 200240 China Tsung-Dao Lee Institute Shanghai 200240 China

The PandaX-Ⅲ experiment will search for neutrinoless double beta decay of 136Xe with high pressure gaseous time projection chambers at the China Jin-Ping underground Laboratory. The tracking feature of gaseous detectors helps suppress the background level, resulting in the improvement of the detection sensitivity. We study a method based on the convolutional neural networks to discriminate double beta decay signals against the background from high energy gammas generated by 214Bi and 2^208 T1 decays based on detailed Monte Carlo simulation. Using the 2-dimensional projections of recorded tracks on two planes, the method successfully suppresses the background level by a factor larger than 100 with a high signal efficiency. An improvement of 62% on the efficiency ratio of Еs/√Еb is achieved in comparison with the baseline in the PandaX-Ⅲ conceptual design report.

关键词： neutrino double beta decay convolutional neural networks background suppression

A feature-wise attention module based on the difference with surrounding features for convolutional neural networks

在线全文

学校读者我要写书评

暂无评论

Frontiers of Computer Science 2023年第6期17卷 77-86页

作者： Shuo TAN Lei ZHANG Xin SHU Zizhou WANG Machine Intelligence Laboratory College of Computer ScienceSichuan UniversityChengdu 610065China

Attention mechanism has become a widely researched method to improve the performance of convolutional neural networks(CNNs).Most of the researches focus on designing channel-wise and spatial-wise attention modules but neglect the importance of unique information on each feature,which is critical for deciding both“what”and“where”to *** this paper,a feature-wise attention module is proposed,which can give each feature of the input feature map an attention ***,the module is based on the well-known surround suppression in the discipline of neuroscience,and it consists of two sub-modules,Minus-Square-Add(MSA)operation and a group of learnable non-linear mapping *** MSA imitates the surround suppression and defines an energy function which can be applied to each feature to measure its *** group of non-linear functions refines the energy calculated by the MSA to more reasonable *** these two sub-modules,feature-wise attention can be well ***,due to the simple structure and few parameters of the two sub-modules,the proposed module can easily be almost integrated into any *** verify the performance and effectiveness of the proposed module,several experiments were conducted on the Cifar10,Cifar100,Cinic10,and Tiny-ImageNet datasets,*** experimental results demonstrate that the proposed module is flexible and effective for CNNs to improve their performance.

关键词： feature-wise attention surround suppression image classification convolutional neural networks

同方期刊数据库

Deep Imitation Learning for Autonomous Vehicles Based on convolutional neural networks

在线全文

学校读者我要写书评

暂无评论

IEEE/CAA Journal of Automatica Sinica 2020年第1期7卷 82-95页

作者： Parham M.Kebria Abbas Khosravi Syed Moshfeq Salaken Saeid Nahavandi Intelligent Systems Research and Innovation(IISRI) Deakin UniversityWaurn PondsVIC 3216Australia

Providing autonomous systems with an effective quantity and quality of information from a desired task is challenging. In particular, autonomous vehicles, must have a reliable vision of their workspace to robustly accomplish driving functions. Speaking of machine vision, deep learning techniques, and specifically convolutional neural networks, have been proven to be the state of the art technology in the field. As these networks typically involve millions of parameters and elements, designing an optimal architecture for deep learning structures is a difficult task which is globally under investigation by researchers. This study experimentally evaluates the impact of three major architectural properties of convolutional networks, including the number of layers, filters, and filter size on their performance. In this study, several models with different properties are developed,equally trained, and then applied to an autonomous car in a realistic simulation environment. A new ensemble approach is also proposed to calculate and update weights for the models regarding their mean squared error values. Based on design properties,performance results are reported and compared for further investigations. Surprisingly, the number of filters itself does not largely affect the performance efficiency. As a result, proper allocation of filters with different kernel sizes through the layers introduces a considerable improvement in the *** of this study will provide the researchers with a clear clue and direction in designing optimal network architectures for deep learning purposes.

关键词： Autonomous vehicles convolutional neural networks deep learning imitation learning

Reconfigurable spatial-parallel stochastic computing for accelerating sparse convolutional neural networks

在线全文

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2023年第6期66卷 267-286页

作者： Zihan XIA Rui WAN Jienan CHEN Runsheng WANG National Key Laboratory of Science and Technology on Communications University of Electronic Science and Technology of China School of Integrated Circuits Peking University

Edge devices play an increasingly important role in the convolutional neural network(CNN)inference. However, the large computation and storage requirements are challenging for resource-and powerconstrained hardware. These limitations might be overcome by exploring the following:(a)error tolerance via approximate computing, such as stochastic computing(SC);(b)data sparsity, including the weight and activation sparsity. Although SC can perform complex calculations with compact and simple arithmetic circuits,traditional SC-based accelerators suffer from the low reconfigurability and long bitstream, further making it difficult to benefit from the data sparsity. In this paper, we propose spatial-parallel stochastic computing(SPSC), which improves the spatial parallelism of the SC-based multiplier to the full extent while consuming fewer logic gates than the fixed-point implementation. Moreover, we present SPA, a highly reconfigurable SPSC-based sparse CNN accelerator with the proposed hybrid zero-skipping scheme(HZSS), to efficiently take advantage of different zero-skipping strategies for different types of layers. Comprehensive experiments show that SPA with up to 2477.6 Gops/W outperforms existing several binary-weight accelerators, SC-based accelerators, and the sparse CNN accelerator considering energy efficiency.

关键词： convolutional neural networks stochastic computing sparse neural networks energy-efficient accelerator high reconfigurability spatial parallelism

同方期刊数据库