检索结果-内蒙古大学图书馆

16th Annual Conference of the International-Speech-Communication-Association (INTERSPEECH 2015)

作者： Yamamoto, Hitoshi Koshinaka, Takafumi NEC Corp Ltd Informat & Media Proc Lab Tokyo Japan

ISBN: (纸本)9781510817906

This paper describes a speaker feature restoration method for improving text-independent speaker recognition with short utterances. The method employs a denoising autoencoder (DAE) to compensate speaker features of a short utterance which contains limited phonetic information. It first estimates phonetic distribution in the utterance as posteriors based on speech models and then transforms an i-vector of the utterance using DAE along with the phonetic posteriors. The DAE-based transformation is able to produce a reliable speaker feature with help of supervised training using pairs of long and short speech segments. Speaker recognition experiments on an NIST SRE task demonstrate a 37.9% error reduction.

关键词： speaker recognition short utterance i-vector feature compensation denoising autoencoder

来源：评论

学校读者我要写书评

暂无评论

denoising-autoencoder-Aided Euclidean Distance Matrix Reconstruction for Connectivity-Based Localization: A Low-Rank Perspective

引用

APPLIED SCIENCES-BASEL 2025年第5期15卷 2656-2656页

作者： Lee, Woong-Hee Ozger, Mustafa Challita, Ursula Song, Taewon Dongguk Univ Seoul Div Elect & Elect Engn Seoul 04620 South Korea Aalborg Univ Dept Elect Syst DK-2450 Copenhagen Denmark Ericsson AB S-16440 Stockholm Sweden Soonchunhyang Univ Dept Internet Things Asan 31538 South Korea

In contrast to conventional localization methods, connectivity-based localization is a promising approach that leverages wireless links among network nodes. Here, the Euclidean distance matrix (EDM) plays a pivotal role in implementing the multidimensional scaling technique for the localization of wireless nodes based on pairwise distance measurements. This is based on the representation of complex datasets in lower-dimensional spaces, resulting from the mathematical property of an EDM being a low-rank matrix. However, EDM data are inevitably susceptible to contamination due to errors such as measurement imperfections, channel dynamics, and clock asynchronization. Motivated by the low-rank property of the EDM, we introduce a new pre-processor for connectivity-based localization, namely denoising-autoencoder-aided EDM reconstruction (DAE-EDMR). The proposed method is based on optimizing the neural network by inputting and outputting vectors of the eigenvalues of the noisy EDM and the original EDM, respectively. The optimized NN denoises the contaminated EDM, leading to an exceptional performance in connectivity-based localization. Additionally, we introduce a relaxed version of DAE-EDMR, i.e., truncated DAE-EDMR (T-DAE-EDMR), which remains operational regardless of variations in the number of nodes between the training and test phases in NN operations. The proposed algorithms show a superior performance in both EDM denoising and localization accuracy. Moreover, the method of T-DAE-EDMR notably requires a minimal number of training datasets compared to that in conventional approaches such as deep learning algorithms. Overall, our proposed algorithms reduce the required training dataset's size by approximately one-tenth while achieving more than twice the effectiveness in EDM denoising, as demonstrated through our experiments.

关键词： connectivity-based localization denoising autoencoder Euclidean distance matrix multidimensional scaling neural networks

来源：评论

学校读者我要写书评

暂无评论

denoising autoencoder for Partial Discharge Identification in Instrument Transformers 31

Denoising Autoencoder for Partial Discharge Identification i...

引用

31st International Conference on Systems, Signals and Image Processing (IWSSIP)

作者： Crivelaro, Matheus Goes Rodrigues, Douglas Gifalli, Andre Papa, Joao Paulo Gonzales, Carlos Guilherme de Souza, Andre Nunes da Silva, Gustavo Vinicius Silveira Neto, Erasmo Sao Paulo State Univ Dept Elect Engn Bauru SP Brazil Sao Paulo State Univ Dept Comp Bauru SP Brazil ISA CTEEP Specialized Maintenance Ctr Bauru SP Brazil High Voltage Equipments HVEX Dept Elect Engn Itajuba Brazil

ISBN: (纸本)9798350391893;9798350391886

Analyzing Partial Discharge (PD) signals is crucial to assessing the health of insulation in high-voltage systems. Nevertheless, noise often distorts these signals, hindering the ability to obtain precise information. This paper proposes a novel deep-learning approach using two denoising autoencoders (DAEs) to learn data representations and eliminate noise during reconstruction. By leveraging DAEs' capacity to capture essential features within the latent space, this method enhances the analysis of PD signals and yields more accurate results. This paper investigates the effectiveness of two deep-learning architectures for denoising partial discharge signals in high-voltage insulation systems. Experimental results carried out on a PD dataset demonstrated the efficiency of the Linear AE model in removing noise in sets A, B, and C suggesting that DAEs hold great promise in PD signal denoising.

关键词： partial discharge signal denoising autoencoder deep learning

来源：评论

学校读者我要写书评

暂无评论

denoising autoencoder Based Delete and Generate Approach for Text Style Transfer 30th

Denoising AutoEncoder Based Delete and Generate Approach for...

引用

30th International Conference on Artificial Neural Networks (ICANN)

作者： Hu, Ting Yang, Haojin Meinel, Christoph Univ Potsdam Hasso Plattner Inst Potsdam Germany

ISBN: (纸本)9783030863654;9783030863647

Text style transfer task is transferring sentences to other styles while preserving the semantics as much as possible. In this work, we study a two-step text style transfer method on non-parallel datasets. In the first step, the style-relevant words are detected and deleted from the sentences in the source style corpus. In the second step, the remaining style-devoid contents are fed into a Natural Language Generation model to produce sentences in the target style. The model consists of a style encoder and a pre-trained denoisingautoencoder. The former extracts style features of each style corpus and the latter reconstructs source sentences during training and generates sentences in the target style during inference from given contents. We conduct experiments on two text sentiment transfer datasets and comprehensive comparisons with other relevant methods in terms of several evaluation aspects. Evaluation results show that our method outperforms others in terms of sentence fluency and achieves a decent tradeoff between content preservation and style transfer intensity. The superior performance on the Caption dataset illustrates our method's potential advantage on occasions of limited data.

关键词： Text style transfer denoising autoencoder Natural Language Generation

来源：评论

学校读者我要写书评

暂无评论

A correlative denoising autoencoder to model social influence for top-N recommender system

引用

Frontiers of Computer Science 2020年第3期14卷 31-43页

作者： Yiteng PAN Fazhi HE Haiping YU School of Computer Science Wuhan UniversityWuhan 430072China

In recent years,there are numerous works been proposed to leverage the techniques of deep learning to improve social-aware recommendation *** most cases,it requires a larger number of data to train a robust deep learning model,which contains a lot of parameters to fit training ***,both data of user ratings and social networks are facing critical sparse problem,which makes it not easy to train a robust deep neural network *** this problem,we propose a novel correlative denoising autoencoder(CoDAE)method by taking correlations between users with multiple roles into account to learn robust representations from sparse inputs of ratings and social networks for *** develop the CoDAE model by utilizing three separated autoencoders to learn user features with roles of rater,truster and trustee,***,on account of that each input unit of user vectors with roles of truster and trustee is corresponding to a particular user,we propose to utilize shared parameters to learn common information of the units that corresponding to same ***,we propose a related regularization term to learn correlations between user features that learnt by the three subnetworks of CoDAE *** further conduct a series of experiments to evaluate the proposed method on two public datasets for Top-N recommendation *** experimental results demonstrate that the proposed model outperforms state-of-the-art algorithms on rank-sensitive metrics of MAP and NDCG.

关键词： social network recommender system denoising autoencoder neural network

来源：评论

学校读者我要写书评

暂无评论

Noise reduction in the spectral domain of hyperspectral images using denoising autoencoder methods

引用

CHEMOMETRICS AND INTELLIGENT LABORATORY SYSTEMS 2020年 203卷 104063-104063页

作者： Zhang, Chu Zhou, Lei Zhao, Yiying Zhu, Susu Liu, Fei He, Yong Zhejiang Univ Coll Biosyst Engn & Food Sci Hangzhou 310058 Peoples R China Minist Agr & Rural Affairs Key Lab Spect Sensing Hangzhou 310058 Peoples R China

denoising of spectra has been a great challenge in hyperspectral image analysis. Near-infrared hyperspectral images of milk powder, rice flour and soybean flour were acquired and denoising in the spectral domain were studied. Noise free spectra and noises were simulated based on sample pixel-wise spectra. The noisy spectra with signal to noise ratio (SNR) around 45 dB (similar to real pixel-wise spectra) were simulated. The simulated noisy spectra were preprocessed by traditional methods as moving average smoothing (MAS), Savitzky-Golay smoothing (SGS), wavelet transform (WT) and empirical mode decomposition (EMD). The basic denoising autoencoder (DAE-1) and the stacked DAE (DAE-2) were studied for denoising. The noisy spectra with SNR around 35 dB and 55 dB were further simulated to explore the effectiveness of DAE based methods. DAE-1 and DAE-2 performed better than the other methods, with higher SNR, lower mean squared error (MSE) and mean absolute error (MAE). The developed DAE methods were applied to real-world pixel-wise spectra with good performances. The overall results proved the feasibility of using DAE based methods for noise reduction in the spectral domain of hyperspectral images, and the DAE based methods have great potential to be extended to spectral denoising of other vibrational spectroscopy techniques.

关键词： denoising autoencoder Noise reduction Hyperspectral image Wavelet transform Spectra simulation Empirical mode decomposition

来源：评论

学校读者我要写书评

暂无评论

Effective network intrusion detection via representation learning: A denoising autoencoder approach

引用

COMPUTER COMMUNICATIONS 2022年 194卷 55-65页

作者： Lopes, Ivandro O. Zou, Deqing Abdulqadder, Ihsan H. Ruambo, Francis A. Yuan, Bin Jin, Hai Huazhong Univ Sci & Technol Sch Comp Sci & Technol Wuhan 430074 Peoples R China Natl Engn Res Ctr Big Data Technol & Syst Wuhan Peoples R China Cluster & Grid Comp Lab Wuhan Peoples R China Serv Comp Technol & Syst Lab Wuhan Peoples R China Nucleo Operac Soc Informacao Praia Cape Verde Big Data Secur Engn Res Ctr Wuhan Peoples R China Huazhong Univ Sci & Technol Sch Cyber Sci & Engn Wuhan 430074 Peoples R China Cihan Univ Erbil Dept Informat & Software Engn Erbil Kurdistan Regio Iraq Shenzhen Huazhong Univ Sci & Technol Res Inst Shenzhen 518057 Peoples R China

The introduction of deep learning techniques in intrusion detection problems has enabled an enhanced standard of detection effectiveness. However, most of the progress has occurred in supervised learning, which required a vast amount of labeled training samples. In the real world, there is a limited amount of labeled data available to train a deep neural network, affecting the classifier's detection performance. Therefore, to address the lack of labeled network traffic required to train an effective supervised classifier, this study introduces a semi -supervised intrusion detection framework that combines the unsupervised and supervised techniques. The unsupervised pre-training approach is implemented based on a denoising autoencoder (DAE), to compress the intrusion dataset and obtain the lower-dimensional features representation. Then a portion of the compressed data is used to train the DNN classifier based on a multiclass supervised approach. The network architecture is optimized by tuning hyper-parameters using a trial-and-error approach. Comparative analysis is performed between the proposed approach and the most relevant deep learning methods available in the literature against the CICIDS2018 dataset, consisting of recent network attack traces. Our approach outperforms competitive methods while maintaining stable classification results above 99.6% on F1-score, precision, and recall metrics. Additionally, it is trained in 64 min while achieving a low false alarm rate. Furthermore, the DAE module reduces the input network traffic data to one-tenth of the size of the input dataset.

关键词： Deep learning denoising autoencoder Intrusion detection Cybersecurity

来源：评论

学校读者我要写书评

暂无评论

Robust and Fast Temperature Extraction for Brillouin Optical Time-Domain Analyzer by Using denoising autoencoder-Based Deep Neural Networks

引用

IEEE SENSORS JOURNAL 2020年第7期20卷 3614-3620页

作者： Wang, Biwei Guo, Nan Wang, Liang Yu, Changyuan Lu, Chao Hong Kong Polytech Univ Dept Elect & Informat Engn Hong Kong Peoples R China Huazhong Univ Sci & Technol Sch Opt & Elect Informat Natl Engn Lab Next Generat Internet Access Syst Wuhan 430074 Peoples R China

A method of robust and fast temperature extraction for Brillouin optical time-domain analyzer (BOTDA) sensing systems using the denoising autoencoder (DAE) based deep neural networks (DNN) is demonstrated. After appropriate training, the DAE suppresses the noise on the measured Brillouin gain spectrum (BGS), and improves the signal-to-noise ratio (SNR) by 9.96dB in our experiment. To extract temperature, the DAE as a basic block is stacked to form the DNN model. Since the DNN model is based on DAE, both denoising and fast temperature extraction can be simultaneously finished using only one DNN model. Thus, it is more robust and faster than the conventional Lorentzian curve fitting (LCF) method, especially when the input data has low SNR level. In the case of 4.6dB SNR, the standard deviation (SD) of the measured temperature at the end of 40km fiber under test (FUT) is reduced from 2.4 degrees C to 1.2 degrees C by using DAE based DNN when compared with that using the LCF method, and the corresponding root-mean-square error (RMSE) is reduced from 2.4 degrees C to 1.3 degrees C. Moreover, since the temperature information can be extracted directly from the experimental BGS data without the need of time-consuming curve fitting and subsequent conversion from Brillouin frequency shift (BFS) to temperature, the speed of temperature extraction using the DAE based DNN is faster by 500 times than that using LCF. Combining the advantages of both denoising and fast processing speed, the DAE based DNN would be a practical way of temperature extraction for the BOTDA systems.

关键词： Brillouin optical time-domain analyzer denoising autoencoder deep neural networks temperature extraction

来源：评论

学校读者我要写书评

暂无评论

Multiscale denoising autoencoder for improvement of target detection

引用

INTERNATIONAL JOURNAL OF REMOTE SENSING 2021年第8期42卷 3002-3016页

作者： Sun, Qiaoqiao Liu, Xuefeng Bourennane, Salah Liu, Bin Aix Marseille Univ Inst Fresnel Cent Marseille CNRS 52 Ave Escadrille Normandie Niemen F-13013 Marseille France Qingdao Univ Sci & Technol Coll Automat & Elect Engn Qingdao Peoples R China

Target detection is one of the most important applications of hyperspectral technology. However, due to spectral variations caused by noise or environment, the within-class variation is enlarged which degrades the performance of detectors, especially when the target size is small. Therefore, improving the detection performance of small targets and noisy targets is a key task. Considering the great feature extraction and representation ability of deep learning models, denoising autoencoder (DAE) is introduced to reconstruct spectrums and exploit the invariant information for target detection. To fully extract the features from the original spectrums, a multiscale denoising autoencoder (MSDAE) model is designed to incorporate complementary informationin in this paper. The final spectrum is fused by reconstructed spectrums from different scales representations, which provides more complex information and more robust features for subsequent spectral identification. Results on simulated hyperspectral images (HSIs) and real-world HSIs demonstrate that the proposed MSDAE model can effectively remove noise interference and lead to great improvements of the target detection. In addition, the proposed method shows significant potential in preserving small targets.

关键词： Target detection hyperspectral image denoising autoencoder spectrum

来源：评论

学校读者我要写书评

暂无评论

Feature learning from incomplete EEG with denoising autoencoder

引用

NEUROCOMPUTING 2015年 165卷 23-31页

作者： Li, Junhua Struzik, Zbigniew Zhang, Liqing Cichocki, Andrzej RIKEN Brain Sci Inst Lab Adv Brain Signal Proc Wako Saitama 3510198 Japan Shanghai Jiao Tong Univ Dept Comp Sci & Engn Key Lab Shanghai Educ Commiss Intelligent Interac Shanghai 200240 Peoples R China Polish Acad Sci Syst Res Inst PL-00901 Warsaw Poland

An alternative pathway for the human brain to communicate with the outside world is by means of a brain computer interface (BCI). A BCI can decode electroencephalogram (EEG) signals of brain activities, and then send a command or an intent to an external interactive device, such as a wheelchair. The effectiveness of the BCI depends on the performance in decoding the EEG. Usually, the EEG is contaminated by different kinds of artefacts (e.g., electromyogram (EMG), background activity), which leads to a low decoding performance. A number of filtering methods can be utilized to remove or weaken the effects of artefacts, but they generally fail when the EEG contains extreme artefacts. In such cases, the most common approach is to discard the whole data segment containing extreme artefacts. This causes the fatal drawback that the BCI cannot output decoding results during that time. In order to solve this problem, we employ the Lomb-Scargle periodogram to estimate the spectral power from incomplete EEG (after removing only parts contaminated by artefacts), and denoising autoencoder (DAE) for learning. The proposed method is evaluated with motor imagery EEG data. The results show that our method can successfully decode incomplete EEG to good effect. (C) 2015 Elsevier B.V. All rights reserved.

关键词： Brain computer interface Spectral power estimation denoising autoencoder Motor imagery Incomplete EEG

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：