检索结果-内蒙古大学图书馆

Unsupervised Anomaly Video Detection via a Double-Flow ConvLSTM Variational autoencoder

IEEE ACCESS 2022年 10卷 44278-44289页

作者： Wang, Lin Tan, Haishu Zhou, Fuqiang Zuo, Wangxia Sun, Pengfei Beihang Univ Sch Instrumentat Sci & Optoelect Engn Beijing 100191 Peoples R China Foshan Univ Sch Phys & Optoelect Engn Foshan 528000 Peoples R China Ji Hua Lab Foshan 528000 Peoples R China Univ South China Sch Elect Engn Hengyang 421001 Peoples R China

With the rapid increase of video surveillance points in the market in recent years, video anomaly detection has gained extensive attention in the security field. At present, the distribution of normal and anomalous data is unbalanced in unlabeled video data. Variational autoencoder (VAE), as one of the typical deep generative models, gets increasingly popular in unsupervised anomaly detection. However, this model is not good at processing time-series data, especially video data. In addition, the strong generalization ability which is over-reconstructing anomaly behavior of many autoencoder-based works leads to the missed anomaly detection. To solve these problems, in this paper, we present a double-flow convolutional long short-term memory variational autoencoder (DF-ConvLSTM-VAE) to model the probabilistic distribution of the normal video in an unsupervised learning scheme, and to reconstruct videos without anomaly objects for anomaly video detection. Experiments verify the effectiveness and competitiveness of our DF-ConvLSTM-VAE on multiple public benchmark datasets. In particular, our model achieves the state-of-the-art performance on anomalous event count.

关键词： Image reconstruction Anomaly detection Data models Feature extraction Mathematical models Unsupervised learning Task analysis autoencoder variational autoencoder LSTM ConvLSTM anomaly detection

来源：评论

学校读者我要写书评

暂无评论

autoencoder-Based Data Sampling for Machine Learning-Based Lithography Hotspot Detection 22

Autoencoder-Based Data Sampling for Machine Learning-Based L...

引用

4th ACM/IEEE Workshop on Machine Learning for CAD (MLCAD)

作者： Ismail, Mohamed Tarek Sharara, Hossam Madkour, Kareem Seddik, Karim Siemens EDA Cairo Egypt Amer Univ Cairo Cairo Egypt

ISBN: (数字)9781665497299

ISBN: (纸本)9781450394864

Technology scaling has increased the complexity of integrated circuit design. It has also led to more challenges in the field of Design for Manufacturing (DFM). One of these challenges is lithography hotspot detection. Hotspots (HS) are design patterns that negatively affect the output yield. Identifying these patterns early in the design phase is crucial for high yield fabrication. Machine Learning-based (ML) hotspot detection techniques are promising since they have shown superior results to other methods such as pattern matching. Training ML models is a challenging task due three main reasons. First, industrial training designs contain millions of unique patterns. It is impractical to train models using this large number of patterns due to limited computational and memory resources. Second, the HS detection problem has an imbalanced nature;datasets typically have a limited number of HS and a large number of non-hotspots. Lastly, hotspot and non-hotspot patterns can have very similar geometries causing models to be susceptible to high false positive rates. Due to these reasons, the use of data sampling techniques is needed to choose the best representative dataset for training. In this paper, a dataset sampling technique based on autoencoders is introduced. The autoencoders are used to identify latent data features that can reconstruct the input patterns. These features are used to group the patterns using Density-based spatial clustering of applications with noise (DBSCAN). Then, the clustered patterns are sampled to reduce the training set size. Experiments on the ICCAD-2019 dataset show that the proposed data sampling approach can reduce the dataset size while maintaining the levels of recall and precision that were obtained using the full dataset.

关键词： autoencoder clustering DBSCAN data reduction design for manufacturing hotspots machine learning sampling

来源：评论

学校读者我要写书评

暂无评论

A Deep Convolutional autoencoder for Automatic Motion Artifact Removal in Electrodermal Activity

引用

IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING 2022年第12期69卷 3601-3611页

作者： Hossain, Md-Billal Posada-Quintero, Hugo F. Chon, Ki H. Univ Connecticut Dept Biomed Engn Storrs CT 06269 USA

Objective: This study aimed to develop a robust and data driven automatic motion artifacts (MA) removal technique from electrodermal activity (EDA) signal. Methods: we proposed a deep convolutional autoencoder (DCAE) approach for automatic MA removal in EDA signals. Our model was trained using several publicly available datasets that were collected using a wide variety of stimuli to cause EDA reactions;the sample size was large (N = 385 subjects). We trained and validated our DCAE network using both Gaussian white noise (GWN) and realistic MA data records collected using a novel circuitry in our lab. We further evaluated and compared the performance of our DCAE model with the existing methods on two independent and unseen datasets called Chon lab motion artifact dataset II (CMAD II) and central nervous system oxygen toxicity dataset (CNS-OT). Results: Our DCAE model showed significantly higher signal-to-noise-power-ratio improvement (SNRimp) and lower mean squared error (MSE) when compared with that of the three previous methods (averaged SNRimp=35.25dB, and MSE =0.028 on the MA-corrupted data). Moreover, the reconstructed EDAs from the CMAD II dataset had a mean correlation value of 0.78 (statistically significantly higher when compared with other methods) with the reference clean data from the motionless hand, whereas the raw MA-corrupted data had a correlation value of only 0.68. Conclusion: The results presented in the paper indicates that our DCAE can remove MAs with higher intensity where the existing methods fails. Significance: Proposed DCAE model can be used to recover a significant amount of otherwise discarded EDA data.

关键词： Electrodermal activity motion artifacts autoencoder convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Dilated projection correction network based on autoencoder for hyperspectral image super-resolution

引用

NEURAL NETWORKS 2022年第0期146卷 107-119页

作者： Wang, Xinya Ma, Jiayi Jiang, Junjun Zhang, Xiao-Ping Wuhan Univ Elect Informat Sch Wuhan 430072 Peoples R China Harbin Inst Technol Sch Comp Sci & Technol Harbin 150001 Peoples R China Ryerson Univ Dept Elect Comp & Biomed Engn Toronto ON M5B 2K3 Canada

This paper focuses on improving the spatial resolution of the hyperspectral image (HSI) by taking the prior information into consideration. In recent years, single HSI super-resolution methods based on deep learning have achieved good performance. However, most of them only simply apply general image super-resolution deep networks to hyperspectral data, thus ignoring some specific characteristics of hyperspectral data itself. In order to make full use of spectral information of the HSI, we transform the HSI SR problem from the image domain into the abundance domain by the dilated projection correction network with an autoencoder, termed as aeDPCN. In particular, we first encode the low-resolution HSI to abundance representation and preserve the spectral information in the decoder network, which could largely reduce the computational complexity. Then, to enhance the spatial resolution of the abundance embedding, we super-resolve the embedding in a coarse-to-fine manner by the dilated projection correction network where the back-projection strategy is introduced to further eliminate spectral distortion. Finally, the predictive images are derived by the same decoder, which increases the stability of our method, even at a large upscaling factor. Extensive experiments on real hyperspectral image scenes demonstrate the superiority of our method over the state-of-the-art, in terms of accuracy and efficiency. (C) 2021 Elsevier Ltd. All rights reserved.

关键词： Hyperspectral image Super-resolution Deep learning autoencoder

来源：评论

学校读者我要写书评

暂无评论

autoencoder Communications with Optimized Interference Suppression for NextG RAN

Autoencoder Communications with Optimized Interference Suppr...

引用

IEEE Future Networks World Forum (FNWF)

作者： Davaslioglu, Kemal Erpek, Tugba Sagduyu, Yalin E. Univ Tech Serv Greenbelt MD 20770 USA Virginia Tech Natl Secur Inst Arlington VA USA

ISBN: (纸本)9781665462501

This paper models an end-to-end communications system for the NextG radio access network (RAN) as an autoencoder (AE) subject to interference effects. The transmitter (coding and modulation) and receiver (demodulation and decoding) are represented as deep neural networks (DNNs) of the encoder and decoder, respectively. The AE communications systems is trained with interference training and randomized smoothing to operate under unknown and dynamic interference (jamming) effects. Compared to conventional communications, the AE communications with interference training and randomized smoothing can achieve up to 36 dB interference suppression with a channel reuse of four for the single antenna case. This paper also extends the AE communications formulation to the multiple-input multiple-output (MIMO) case under interference effects and shows the bit error rate (BER) improvement compared to conventional MIMO communications.

关键词： autoencoder deep learning interference suppression SISO MIMO

来源：评论

学校读者我要写书评

暂无评论

autoencoder With Emotion Embedding for Speech Emotion Recognition

引用

IEEE ACCESS 2021年 9卷 51231-51241页

作者： Zhang, Chenghao Xue, Lei Shanghai Univ Sch Commun & Informat Engn Shanghai 200444 Peoples R China

An important part of the human-computer interaction process is speech emotion recognition (SER), which has been receiving more attention in recent years. However, although a wide diversity of methods has been proposed in SER, these approaches still cannot improve the performance. A key issue in the low performance of the SER system is how to effectively extract emotion-oriented features. In this paper, we propose a novel algorithm, an autoencoder with emotion embedding, to extract deep emotion features. Unlike many previous works, instance normalization, which is a common technique in the style transfer field, is introduced into our model rather than batch normalization. Furthermore, the emotion embedding path in our method can lead the autoencoder to efficiently learn a priori knowledge from the label. It can enable the model to distinguish which features are most related to human emotion. We concatenate the latent representation learned by the autoencoder and acoustic features obtained by the openSMILE toolkit. Finally, the concatenated feature vector is utilized for emotion classification. To improve the generalization of our method, a simple data augmentation approach is applied. Two publicly available and highly popular databases, IEMOCAP and EMODB, are chosen to evaluate our method. Experimental results demonstrate that the proposed model achieves significant performance improvement compared to other speech emotion recognition systems.

关键词： Feature extraction Speech recognition Emotion recognition Spectrogram Noise reduction Hidden Markov models Acoustics Speech emotion recognition autoencoder emotion embedding instance normalization

来源：评论

学校读者我要写书评

暂无评论

Survival stratification for colorectal cancer via multi-omics integration using an autoencoder-based model

引用

EXPERIMENTAL BIOLOGY AND MEDICINE 2022年第11期247卷 898-909页

作者： Song, Hu Ruan, Chengwei Xu, Yixin Xu, Teng Fan, Ruizhi Jiang, Tao Cao, Meng Song, Jun Xuzhou Med Univ Dept Gastrointestinal Surg Affiliated Hosp Xuzhou 221002 Jiangsu Peoples R China Xuzhou Med Univ Dept Anorectal Surg Affiliated Hosp Xuzhou 221002 Jiangsu Peoples R China

Prognosis stratification in colorectal cancer helps to address cancer heterogeneity and contributes to the improvement of tailored treatments for colorectal cancer patients. In this study, an autoencoder-based model was implemented to predict the prognosis of colorectal cancer via the integration of multi-omics data. DNA methylation, RNA-seq, and miRNA-seq data from The Cancer Genome Atlas (TCGA) database were integrated as input for the autoencoder, and 175 transformed features were produced. The survival-related features were used to cluster the samples using k-means clustering. The autoencoder-based strategy was compared to the principal component analysis (PCA)-, t-distributed random neighbor embedded (t-SNE)-, non-negative matrix factorization (NMF)-, or individual Cox proportional hazards (Cox-PH)-based strategies. Using the 175 transformed features, tumor samples were clustered into two groups (G1 and G2) with significantly different survival rates. The autoencoder-based strategy performed better at identifying survival-related features than the other transformation strategies. Further, the two survival groups were robustly validated using "hold-out" validation and five validation cohorts. Gene expression profiles, miRNA profiles, DNA methylation, and signaling pathway profiles varied from the poor prognosis group (G2) to the good prognosis group (G1). miRNA-mRNA networks were constructed using six differentially expressed miRNAs (let-7c, mir-34c, mir-133b, let-7e, mir-144, and mir-106a) and 19 predicted target genes. The autoencoder-based computational framework could distinguish good prognosis samples from bad prognosis samples and facilitate a better understanding of the molecular biology of colorectal cancer.

关键词： autoencoder deep learning K-means clustering multi-omics survival

来源：评论

学校读者我要写书评

暂无评论

A Bi-LSTM and autoencoder Based Framework for Multi-step Flight Trajectory Prediction 8

A Bi-LSTM and AutoEncoder Based Framework for Multi-step Fli...

引用

8th International Conference on Control and Robotics Engineering (ICCRE)

作者： Wu, Han Liang, Yan Zhou, Bin Sun, Hao Northwestern Polytech Univ Sch Automat Xian Peoples R China

ISBN: (纸本)9798350345650

Trajectory prediction (TP) is a key component in the route planning for civil aircraft. Most existing methods obtain multi-step TP via iterating the one-step TP model, which generally generates large cumulative error due to deviate from the original evolutionary pattern. To improve the situation, this paper proposes a multi-step TP framework with three modules: the Bi-directional Long Short-Term Memory Network (Bi-LSTM) based multi-step TP module, autoencoder based multi-step TP module, and voting fusion module. In the Bi-LSTM based multistep TP method, to avoid the forgetting of evolutionary characteristics, the Bi-LSTM is designed to directly extract the mapping relationship between input of historical trajectory fragments and output of multi-step labels via data- driven method. In the autoencoder based multi-step TP module, the Bi-LSTM is deigned to learn mapping relationship between the input and core evolutionary features from output labels extracted via the encoder, and then the decoder is adopted to reconstruct predictions by outputs from Bi-LSTM. Third, the voting method was used to fuse the per-dimension predictions from the above two modules and further to refine multi-step predictions. The proposed multi-step TP framework is applied to real flight trajectory prediction of civil aircraft and outperforms multiple deep learning methods in the terms of RMSE and MAE.

关键词： multi-step trajectory prediction autoencoder BiLSTM time series analysis

来源：评论

学校读者我要写书评

暂无评论

CηιDAE: Cryptographically Distinguishing autoencoder for Cipher Cryptanalysis

CηιDAE: Cryptographically Distinguishing Autoencoder for C...

引用

IEEE Conference on Global Communications (IEEE GLOBECOM) - Intelligent Communications for Shared Prosperity

作者： Phan, Raphael C. -W. Pal, Arghya Wong, KokSheik Rajanala, Sailaja Monash Univ Sch IT Psi Phi Lab Malaysia Campus Subang Jaya Malaysia

ISBN: (纸本)9798350310900

We propose a new autoencoder (AE) construction C eta iota DAE (Cryptographically Distinguishing AE) based on a novel loss formulation to solve the cipher cryptanalysis distinguishing problem in the domain of cryptology. Vanilla AE and variational AE are unable to address this problem as they are designed to draw new samples which are either similar to the input sample or are from the same distribution. Such generated samples do not facilitate the cryptanalysis task. We show that our AE construction enables the discovery of cipher distinguishers, which are the fundamental building blocks that make or break new cipher design proposals. This also answers an open question on the applicability of autoencoders for cipher cryptanalysis;as to date, only discriminative models have been applied for cryptanalysis problems. To the best of our knowledge, C eta iota DAE is the first-known generative model designed to solve cryptanalysis problems. We apply our C eta iota DAE model to discover distinguishing properties for up to 10 rounds of the NSA-designed Speck32/64 cipher that allows to distinguish it from a random permutation. This contrasts with the best-known machine learning-discovered neural distinguisher in the literature that covers up to 8 rounds of Speck32/64. Unlike these recent related work which leverage on white box analysis and human-guided differential or linear analysis in order for machine learning models to be applicable, our C eta iota DAE distinguisher does not require prior human cryptanalytic knowledge. This motivates the new direction of human-unsupervised machine learning-based cryptanalysis techniques.

关键词： Generative model autoencoder distinguisher cryptanalysis cipher

来源：评论

学校读者我要写书评

暂无评论

SEMI-SUPERVISED autoencoder WITH JOINT LOSS LEARNING FOR BEARING FAULT DETECTION

SEMI-SUPERVISED AUTOENCODER WITH JOINT LOSS LEARNING FOR BEA...

引用

ASME International Design Engineering Technical Conferences / 43rd Annual Computers and Information in Engineering Conference (IDETC-CIE) / 35th Conference on Mechanical Vibration and Sound (VIB)

作者： Zhou, Kai Zhang, Yang Tang, Jiong Michigan Technol Univ Dept Mech Engn Mech Houghton MI 49931 USA Hong Kong Polytech Univ Dept Civil & Environm Engn Hong Kong Peoples R China Univ Connecticut Dept Mech Engn Storrs CT 06269 USA Univ Connecticut Dept Mech Engn Storrs CT 06269 USA

ISBN: (纸本)9780791887400

Timely and accurate bearing fault detection plays an important role in various industries. Data-driven deep learning methods have recently become a prevailing approach for bearing fault detection. Despite the success of deep learning, fault diagnosis performance is hinged upon the size of labeled data, the acquisition of which oftentimes is expensive in actual practice. Unlabeled data, on the other hand, are inexpensive. To fully utilize a large amount of unlabeled data together with limited labeled data to enhance fault detection performance, in this research, we develop a semi-supervised learning method built upon the autoencoder. In this method, a joint loss is established to account for the effects of both the labeled and unlabeled data, which is subsequently used to direct the backpropagation training. Systematic case studies using the Case Western Reserve University (CWRU) rolling bearing dataset are carried out, in which the effectiveness of this new method is verified by comparing it with other benchmark models.

关键词： Rolling bearing deep learning fault detection semi-supervised learning autoencoder joint loss

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：