检索结果-内蒙古大学图书馆

Learning Facial Expression Codes with sparse auto-encoder

2nd International Conference on Mechatronics and Control Engineering (ICMCE 2013)

作者： Hu, Dekun Duan, Guiduo Chengdu Univ Inst Higher Educ Sichuan Prov Key Lab Pattern Recognit & Intelligent Informat P Chengdu 610106 Peoples R China Univ Elect Sci & Technol China Sch Comp Sci & Engn Chengdu 610054 Peoples R China

ISBN: (纸本)9783037858943

A sparse auto-encoder model was trained to extract the code of different facial expression, which comprises four encoder layers and three decode layers, the representation locating in the fourth layer (code layer) is the features expected. With large amounts of patches randomly selected from training faces, the model was trained firstly via backpropagation which minimizes an unsupervised sparse reconstruction error, and then a softmax classifier was learned for supervised classification. The input vector for the classification is the feature of facial image induced by the learned sparse auto-encoder and two key operations (convolving and pooling). Using a small number of hidden units per layer and a relatively small number of training set, the proposed model achieves excellent performance in the experiments.

关键词： Feature learning facial expression sparse auto-encoder

来源：评论

学校读者我要写书评

暂无评论

Classification of medical images based on deep stacked patched auto-encoders

引用

MULTIMEDIA TOOLS AND APPLICATIONS 2020年第35-36期79卷 25237-25257页

作者： Ben Ali, Ramzi Ejbali, Ridha Zaied, Mourad Univ Gabes ENIG Res Team Intelligent Machines RTIM Zrig 6029 Gabes Tunisia RTIM Gabes Tunisia

The concept of artificial intelligence is not new. Without going into details of the evolution of artificial intelligence, we can confess that recent techniques of deep neural networks have considerably relaunched the trend with a significant advance namely the ability to automatically learn high-level concepts. However, a great step has been taken in deep learning to help researchers perform segmentation, feature extraction, classification and detection from raw medical images. This paper concerns the automatic classification of medical images with deep neural networks. We aimed at developing a system for automatic classification of medical images and detection of anomalies in order to provide a decision-making tool for the doctor. In this component we proposed a method for classifying medical images based on deep neural network using sparse coding and wavelet analysis. Serval real databases are used to test the proposed methods: MIAS and DDSM for mammogram images, LIDC-IDRI for lung images and dental dataset images. Classifications rates given by our approach show a clear improvement compared to those cited in this article.

关键词： Images classification Pattern recognition Deep learning Wavelet network Deep neural wavelets network sparse auto-encoder

来源：评论

学校读者我要写书评

暂无评论

A Novel Network Intrusion Detection Method Based on DSAE-PSOCNN Model

引用

ENGINEERING LETTERS 2022年第4期30卷 1306页

作者： Sun, Qiaochu Dai, Hong Xu, Yao Shi, Tianwei Univ Sci & Technol Liaoning Sch Comp Sci & Software Engn Anshan Peoples R China

Network intrusion detection plays a vital role in information network security protection. To solve the deficiency of the feature dimension reduction and the detection performance, we propose a novel intrusion detection model, referred to as DSAE-PSOCNN. The proposed model is a deep learning model which fuses with the improved sparse auto-encoder (SAE) and optimization of Convolutional Neural Network (CNN) based on Particle Swarm Optimization (PSO). The intrusion data has problems such as high feature dimension and noise data, which may cause overfitting. The model of DSAE designs SAE based on Directed Acyclic Graph (DAG) structure to solve the above problems. It extracts differentially the characteristics of each attack type by analytic hierarchy process and gets the high correlation features. The hyper-parameters of CNN are optimized by using PSO algorithm in PSOCNN model to select independently the best CNN structure without the guidance of experience. Finally, the excellence of the proposed model is verified on the CIC-IDS2017 dataset. DSAE-PSOCNN model achieves an accuracy of 98.6% and it compares with the other three models. We conclude that DSAE-PSOCNN model outperforms the comparative models in the precision and recall rate. The suggested model provides an effective solution to the feature dimension reduction.

关键词： intrusion detection sparse auto-encoder convolutional neural network particle swarm optimization directed acyclic graph

来源：评论

学校读者我要写书评

暂无评论

WiFi-Based Driver's Activity Monitoring with Efficient Computation of Radio-Image Features

引用

SENSORS 2020年第5期20卷 1381-1381页

作者： Akhtar, Zain Ul Abiden Wang, Hongyu Dalian Univ Technol Fac Elect Informat & Elect Engn Dalian 116023 Peoples R China

Driver distraction and fatigue are among the leading contributing factors in various fatal accidents. Driver activity monitoring can effectively reduce the number of roadway accidents. Besides the traditional methods that rely on camera or wearable devices, wireless technology for driver's activity monitoring has emerged with remarkable attention. With substantial progress in WiFi-based device-free localization and activity recognition, radio-image features have achieved better recognition performance using the proficiency of image descriptors. The major drawback of image features is computational complexity, which increases exponentially, with the growth of irrelevant information in an image. It is still unresolved how to choose appropriate radio-image features to alleviate the expensive computational burden. This paper explores a computational efficient wireless technique that could recognize the attentive and inattentive status of a driver leveraging Channel State Information (CSI) of WiFi signals. In this novel research work, we demonstrate an efficient scheme to extract the representative features from the discriminant components of radio-images to reduce the computational cost with significant improvement in recognition accuracy. Specifically, we addressed the problem of the computational burden by efficacious use of Gabor filters with gray level statistical features. The presented low-cost solution requires neither sophisticated camera support to capture images nor any special hardware to carry with the user. This novel framework is evaluated in terms of activity recognition accuracy. To ensure the reliability of the suggested scheme, we analyzed the results by adopting different evaluation metrics. Experimental results show that the presented prototype outperforms the traditional methods with an average recognition accuracy of 93.1% in promising application scenarios. This ubiquitous model leads to improve the system performance significantly for the diverse

关键词： Channel State Information (CSI) wireless sensing activity recognition driver distraction efficient computation sparse auto-encoder

来源：评论

学校读者我要写书评

暂无评论

Heterogeneous transfer learning based on Stack sparse auto-encoders for fault diagnosis

Heterogeneous transfer learning based on Stack Sparse Auto-E...

引用

Chinese automation Congress (CAC)

作者： Wang Chunfeng Lv Zheng Zhao Jun Wang Wei Dalian Univ Technol Sch Control Sci & Technol Dalian Peoples R China

ISBN: (纸本)9781728113128

Fault diagnosis can reduce the risk of accidental failure and play a vital role in ensuring the reliability and safety of industrial systems. The traditional fault diagnosis algorithms mostly require enough training samples. However, in many cases it can be difficult and expensive in some scenarios. In this paper, the auxiliary domain data are used to train the learner and a novel heterogeneous transfer learning method is proposed for fault diagnosis. Data from source domain and target domain are represented by heterogeneous characteristics of different dimensions in heterogeneous transfer learning. We project the source domain and the target domain into the same feature space through two different auto-encoders. Then the similarity of distribution between source domain and target domain could be evaluated. The concept of distance to the center of the domain is introduced to evaluate the similarity of distribution between source domain and target domain. Firstly, it is introduced into the projection process using a small number of target domain labeled samples supervised training sparse auto-encoders (SAEs). Then, the second encoder is used to extract further features. Finally, the source domain data was used to train SVM, and use it to diagnose the target domain data. The experiment result shows that classifier trained by different auxiliary domain data have different performance for target data. The proposed approach performs better than the traditional machine learning approach when there is little labelled data in the target domain.

关键词： Heterogeneous transfer learning sparse auto-encoder distance to the center of the domain fault diagnosis

来源：评论

学校读者我要写书评

暂无评论

automatic Modulation Classification for Short Burst Underwater Acoustic Communication Signals Based on Hybrid Neural Networks

引用

IEEE ACCESS 2020年 8卷 227793-227809页

作者： Li, Yongbin Wang, Bin Shao, Gaoping Shao, Shuai PLA Strateg Support Force Informat Engn Univ Zhengzhou 450001 Peoples R China

automatic modulation classification (AMC) is challenging for short burst underwater acoustic (UWA) communication signals. Difficulties include but are not limited to the poor UWA channels, impulsive noise, and data scarcity. To address these problems, a method based on hybrid neural networks (HNNs) is proposed in this paper. First, an impulsive noise preprocessor is adopted to mitigate the impulse in the received signals. Subsequently, an HNN consisting of an attention aided convolutional neural network (Att-CNN) and a sparse auto-encoder is built to extract features from the temporal waveforms and square spectra of the preprocessed signals after burst detection. Finally, a late fusion is made to combine the prediction results of the two sub-networks. To overcome the variable signal duration relative to the fixed input size of the Att-CNN, a data-reusing approach is proposed to perform dimension preprocessing on the waveforms. Moreover, a transfer learning strategy is introduced to resolve the issue of insufficient training data from the testing channel. The results of simulation experiments and practical signal tests both demonstrate that the proposed method is robust against UWA channels and ambient noise. Our approach significantly outperforms existing deep learning-based methods in dealing with short and weak signal bursts, while requiring less training data from the testing channel.

关键词： Testing Feature extraction Training data Phase shift keying Modulation Frequency shift keying Data models automatic modulation classification hybrid neural networks attention aided convolutional neural networks sparse auto-encoder data-reusing transfer learning

来源：评论

学校读者我要写书评

暂无评论

Low-Resolution Person Re-identification by a Discriminative Resolution-Invariant Network 14th

Low-Resolution Person Re-identification by a Discriminative ...

引用

14th Chinese Conference on Biometric Recognition (CCBR)

作者： Guo, Tongtong Lai, Jianhuang Feng, Zhanxiang Chen, Zeyu Xie, Xiaohua Zheng, Weishi Sun Yat Sen Univ Sch Data & Comp Sci Guangzhou Peoples R China Sun Yat Sen Univ Sch Elect & Informat Technol Guangzhou Peoples R China Guangdong Key Lab Informat Secur Technol Guangzhou Peoples R China Minist Educ Key Lab Machine Intelligence & Adv Comp Guangzhou Peoples R China

ISBN: (纸本)9783030314569;9783030314552

Person Re-identification (re-id) needs to tackle with the problem of changing resolutions because the pedestrians from surveillance systems or public datasets have low-resolution problem (LR-REID) including low quality, blurry textures and so on, which results in a difficult challenge to extract the identity information under various resolutions. However, most existing re-id models are trained by high-resolution (HR) images, which will achieve poor performance when conducted directly on low-resolution images. In this paper, we propose a novel Discriminative Resolution-invariant Network (DRINet) to explore the subspace where LR and HR features are highly correlated and we can extract discriminant features in the commonly shared feature space. Firstly, we adopt ResNet as the backbone and impose the softmax loss together with the triplet loss to learn distinguishing features. Secondly, we impose the KL divergence loss on the backbone features to minimize the discrepancies between LR and HR features. Finally, we integrate the sparse auto-encoder (SAE) structure to find a subspace which is robust to the resolution variations. Experimental results verify the effectiveness of the DRINet in improving the LR-REID performance and the superiority of the DRINet against the state-of-the-art methods.

关键词： Person Re-identification Low resolution sparse auto-encoder KL divergence

来源：评论

学校读者我要写书评

暂无评论

Extending an Anomaly Detection Benchmark with auto-encoders, Isolation Forests, and RBMs 25th

Extending an Anomaly Detection Benchmark with Auto-encoders,...

引用

25th International Conference on Information and Software Technologies (ICIST)

作者： Pijnenburg, Mark Kowalczyk, Wojtek Leiden Univ Leiden Inst Adv Comp Sci Niels Bohrweg 1 Leiden Netherlands Netherlands Tax & Customs Adm Utrecht Netherlands

ISBN: (纸本)9783030302757;9783030302740

In this paper, the recently published benchmark of Goldstein and Uchida [3] for unsupervised anomaly detection is extended with three anomaly detection techniques: sparse auto-encoders, Isolation Forests, and Restricted Boltzmann Machines. The underlying mechanisms of these algorithms differ substantially from the more traditional anomaly detection algorithms, currently present in the benchmark. Results show that in three of the ten data sets, the new algorithms surpass the present collection of 19 algorithms. Moreover, a relation is noted between the nature of the outliers in a data set and the performance of specific (clusters of) anomaly detection algorithms.

关键词： Anomaly detection sparse auto-encoder Isolation forest Restricted Boltzmann machine Benchmark U-index

来源：评论

学校读者我要写书评

暂无评论

Classifying aircraft based on sparse recovery and deep-learning

引用

JOURNAL OF ENGINEERING-JOE 2019年第21期2019卷 7464-7468页

作者： Wang Wenying Wei Yao Zhen Xuanxuan Yu Hui Wang Ruqi Nanjing Res Inst Elect Technol Nanjing 210039 Jiangsu Peoples R China

A hybrid CS-DL method for aircraft classification in complex electromagnetic environment is introduced. To classify aircraft from interfered radar echoes, the authors propose a novel classification method based on compressed sensing (CS) and deep-learning (DL). After recovering the spectrum polluted by jamming signals by using CS, they exploit sparse auto-encoder (SAE) to extract modulation features and then classify aircraft. The method is tested by 536 flights of three types of airplanes, and the results show that the correct classification rate reaches 75% even when 41% of the pulses are interfered.

关键词： jamming signal classification learning (artificial intelligence) feature extraction neural nets radar cross-sections aerospace computing interference (signal) deep-learning sparse auto-encoder correct classification rate sparse recovery hybrid CS-DL aircraft classification complex electromagnetic environment interfered radar echoes novel classification method compressed sensing jamming signals modulation feature extraction

来源：评论

学校读者我要写书评

暂无评论

Turn-to-Turn Short Circuit of Motor Stator Fault Diagnosis Using Dropout Rate Improved Deep sparse autoencoder 3

Turn-to-Turn Short Circuit of Motor Stator Fault Diagnosis U...

引用

3rd IEEE Advanced Information Technology, Electronic and automation Control Conference (IAEAC)

作者： Wang, Botao Ma, Baohui Xu, Kexing Zheng, Tingting Xi An Jiao Tong Univ Sch Elect Engn Xian 710049 Shaanxi Peoples R China State Key Lab Large Elect Drive Syst & Equipment Tianshui 741020 Gansu Peoples R China

ISBN: (纸本)9781538645093

Motor is one of the most frequently used machines in industry. Ensuring the reliability of motor and timely identifying the fault type of motor can guarantee the properly working and prevent the great loss. In this paper, a turn-to-turn short circuit of motor stator and unbalance power supply fault diagnosis system based on Deep sparse auto-encoder and Soft-max Classifier is proposed. The selection of neural network parameters and their influence are systematic given. A new method where dropout rates of sparse auto-encoder are various in each layer is adopted to increase the stability and convergence speed of training process. Finally, the proposed system is applied to an experiment on a motor in laboratory. The conclusion shows the ability to identify the fault type of motor at the continuous state that the accuracy reaches 100%, when only the data from motor at discrete state point are used in training, which makes the system extensible and promising.

关键词： Fault Diagnosis sparse auto-encoder Soft-max Classifier Parameter Selection and Optimization

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：