检索结果-内蒙古大学图书馆

Unsupervised Pre-Training of Imbalanced Data for Identification of Wafer Map Defect Patterns

IEEE ACCESS 2021年 9卷 52352-52363页

作者： Shon, Ho Sun Batbaatar, Erdenebileg Cho, Wan-Sup Choi, Seong Gon Chungbuk Natl Univ Res Inst Comp & Informat Commun Cheongju 28644 South Korea Chungbuk Natl Univ Sch Elect Comp Engn Cheongju 28644 South Korea Chungbuk Natl Univ Dept Management Informat Syst Cheongju 28644 South Korea Chungbuk Natl Univ Sch Informat & Commun Engn Cheongju 28644 South Korea

Visual defect inspection and classification are significant steps of most manufacturing processes in the semiconductor and electronics industries. Known and unknown defects on wafer maps tend to cluster, and these spatial patterns provide valuable process information for supporting manufacturing in determining the root causes of abnormal processes. In previous studies, data augmentation-based deep learning (DL) techniques were most commonly used for the identification of wafer map defect patterns (WMDP). Data augmentation is an effective technique for improving the accuracy of modern image classifiers. However, current data augmentation implementations were manually designed for the WMDP problem. In this study, we propose a DL-based method with automatic data augmentation for the WMDP task. Basically, it focuses on learning effective discriminative features, from wafer maps, through a deep network structure. The network consists of a convolution-based variational autoencoder (CVAE) sequentially. First, we pre-trained the CVAE on large training data in an unsupervised manner. Second, we fine-tuned the encoder of the CVAE, which was followed by a neural network (NN) classifier, in a supervised manner. Additionally, we describe a simple procedure for automatically searching for improved data augmentation policies. The policy mainly consists of five image processing functions: rotation, flipping, shifting, shearing range, and zooming. The effectiveness of the proposed method was demonstrated through experimental results obtained from a simulation dataset and a real-world wafer map dataset (WM-811K). This study provides guidance for the application of deep learning in semiconductor manufacturing processes to improve product quality and yield.

关键词： Feature extraction Fabrication Semiconductor device modeling Deep learning Visualization convolutional codes Training Classification convolutional variational autoencoder deep learning imbalanced data neural network unsupervised pre-training variational autoencoder wafer map defect patterns

来源：评论

学校读者我要写书评

暂无评论

Comparison of an Automatic Classification of Partial Dischage Patterns for Large Hydrogenerator

Comparison of an Automatic Classification of Partial Dischag...

引用

IEEE International Conference on Prognostics and Health Management (ICPHM)

作者： Kokoko, Olivier Hudon, Claude Levesque, Melanie Amyot, Normand Zemouri, Ryad IREQ Hydro Quebec Varennes PQ Canada 1IREQ Hydro Quebec Varennes PQ Canada CNAM CEDRIC Lab Paris France

ISBN: (纸本)9781665419703

More and more scientific disciplines are using deep techniques for the automatic classification of massive high dimensionality unlabeled data. Among these disciplines, the classification of partial discharge (PD) patterns is one that represents major challenges in the field of hydrogenerator diagnosis. This paper proposes a method of comparison of five classification topology based on a single convolutional variational autoencoder (CVAE) and ten classifiers. The comparison is based on five cases exploiting all the same database, but using five different feature extraction rules to create the input vectors of the neural networks. These feature extraction rules are based on the expert judgement and are automatically computed in the preprocessing stage. Analysis of the output of all classifiers for each topology suggests that the accuracy level of the classification can be significantly improved by refining the feature extraction rules. Moreover, the visualization of the 2D latent space from the CVAE also suggests that the accuracy level can be even further improved if the whole dataset is considered instead of a smaller reference dataset randomly selected. Results raise many questions about the performance of feature extraction rules and the possibilities to better handling classification of large databases such as the one of PD measurement files used for hydrogenerator diagnosis.

关键词： convolutional variational autoencoder diagnosis partial discharges deep neural networks expert knowledge feature extraction hydro-generators latent space partial discharge pattern

来源：评论

学校读者我要写书评

暂无评论

Learning Latent Representation of Freeway Traffic Situations from Occupancy Grid Pictures Using variational autoencoder

引用

ENERGIES 2021年第17期14卷 5232-5232页

作者： Rakos, Oliver Becsi, Tamas Aradi, Szilard Gaspar, Peter Budapest Univ Technol & Econ Dept Control Transportat & Vehicle Syst H-1111 Budapest Hungary Inst Comp Sci & Control Syst & Control Lab H-1111 Budapest Hungary

Several problems can be encountered in the design of autonomous vehicles. Their software is organized into three main layers: perception, planning, and actuation. The planning layer deals with the sort and long-term situation prediction, which are crucial for intelligent vehicles. Whatever method is used to make forecasts, vehicles' dynamic environment must be processed for accurate long-term forecasting. In the present article, a method is proposed to preprocess the dynamic environment in a freeway traffic situation. The method uses the structured data of surrounding vehicles and transforms it to an occupancy grid which a convolutional variational autoencoder (CVAE) processes. The grids (2048 pixels) are compressed to a 64-dimensional latent vector by the encoder and reconstructed by the decoder. The output pixel intensities are interpreted as probabilities of the corresponding field is occupied by a vehicle. This method's benefit is to preprocess the structured data of the dynamic environment and represent it in a lower-dimensional vector that can be used in any further tasks built on it. This representation is not handmade or heuristic but extracted from the database patterns in an unsupervised way.

关键词： NGSIM occupancy grid convolutional variational autoencoder

来源：评论

学校读者我要写书评

暂无评论

Deep generative model for probabilistic wind speed and wind power estimation at a wind farm

引用

ENERGY SCIENCE & ENGINEERING 2022年第6期10卷 1855-1873页

作者： Salazar, Andres A. Zheng, Jiafeng Che, Yuzhang Xiao, Feng Tokyo Inst Technol Dept Mech Engn Tokyo Japan Chengdu Univ Informat Technol Coll Atmospher Sci Chengdu 610225 Peoples R China

This work introduces a novel method to generate probabilistic hub-height wind speed forecasts aimed at power output prediction. We employ state-of-the-art convolutional variational autoencoders (CVAEs) trained with historical wind speed observations, multivariable outputs (wind speed, direction, temperature, pressure, and humidity) from a numerical weather prediction (NWP) model and spatio-temporal encodings. After training, we exploit the CVAE data generating capabilities to produce probabilistic forecasts from the same deterministic dynamical NWP model. The resulting probabilistic forecast provides an insight into the uncertainty of the original deterministic input that compensate for errors due to numerical discretization, inaccuracies in initial/boundary conditions and parameterizations and, most importantly wind speed fluctuations due to complex terrain features. To show the performance of the proposed model, we validate the approach with forecasted and observed data for fifteen sites in a wind farm in Awaji Island, Japan, in a challenging zone with complex topography and highly fluctuating wind patterns. We show that the proposed method provides improved wind speed forecasts from both deterministic (reduced root mean square error) and probabilistic (reduced continuous ranked probability score) standpoints. We also use ensemble forecast validation methods to assess the statistical properties of the CVAE and conclude that in zones with rapidly changing wind speed dynamics, the CVAE ensemble performs in a superior way when compared to the analog ensemble method. Finally, we also provide an insight on how the results can be used to obtain reliable wind power estimations.

关键词： convolutional variational autoencoder generative model hub-height wind probabilistic forecasting wind power

来源：评论

学校读者我要写书评

暂无评论

Integration of Deep Learning-Based Inversion and Upscaled Mass-Transfer Model for DNAPL Mass-Discharge Estimation and Uncertainty Assessment

引用

WATER RESOURCES RESEARCH 2022年第10期58卷 e2022WR033277-e2022WR033277页

作者： Kang, Xueyuan Kokkinaki, Amalia Shi, Xiaoqing Yoon, Hongkyu Lee, Jonghyun Kitanidis, Peter K. Wu, Jichun Nanjing Univ Sch Earth Sci & Engn Minist Educ Key Lab Suficial Geochem Nanjing Peoples R China Univ San Francisco Dept Environm Sci San Francisco CA USA Sandia Natl Labs Geosci Res & Applicat POB 5800 Albuquerque NM 87185 USA Univ Hawaii Manoa Dept Civil & Environm Engn Honolulu HI 96822 USA Univ Hawaii Manoa Water Resources Res Ctr Honolulu HI 96822 USA Stanford Univ Dept Civil & Environm Engn Stanford CA 94305 USA

The challenges posed by high-resolution characterization of dense nonaqueous phase liquid (DNAPL) source zone architecture (SZA) have motivated the development of simpler upscaled models that rely on domain-averaged metrics to capture the average mass discharge downstream of source zones (SZ). However, SZA is highly irregular, making estimation of these domain-averaged metrics from sparse borehole data extremely difficult. Poor estimation of SZ metrics means that upscaled models cannot reproduce the multistage effluent concentrations. Bayesian inversion methods can be used to obtain accurate estimates of SZ metrics and their uncertainties from sparse data, and from there, upscaled models can better reproduce multistage effluent concentrations. This work presents a framework for integrating a deep-learning-based 3D SZA inversion method named convolutional variational autoencoder-Ensemble Smoother with Multiple Data Assimilation (CVAE-ESMDA) with a process-based (PB) upscaled mass-transfer model. This framework can utilize sparse SZ data to estimate directly mass discharge without multiphase modeling. First, CVAE-ESMDA estimates the SZA by conditioning on sparse data, which is then used as input in the upscaled model for mass-discharge estimation. We evaluated our framework on two real and 30 synthetic bench-scale experiments, with significantly different SZAs and multistage effluent concentrations. The results demonstrate that the CVAE-based inversion method captures the temporal variations in SZ metrics better than standard ordinary kriging. With the improved SZ metrics, the PB model more accurately reproduces the salient patterns of the multistage mass-discharge profiles and associated uncertainty. This approach can be used to provide valuable input for risk-based decision making in remediation applications.

关键词： DNAPL mass discharge 3D DNAPL characterization convolutional variational autoencoder upscaled modeling multistage dissolution

来源：评论

学校读者我要写书评

暂无评论

Characterization of the non-Gaussian hydraulic conductivity field via deep learning-based inversion of hydraulic-head and self-potential data

引用

JOURNAL OF HYDROLOGY 2022年 610卷

作者： Han, Zheng Kang, Xueyuan Wu, Jichun Shi, Xiaoqing Nanjing Univ Sch Earth Sci & Engn Key Lab Surficial Geochem Minist Educ Nanjing 210023 Peoples R China

Accurate characterization of the spatial heterogeneity of hydraulic properties such as hydraulic conductivity (K) is essential for understanding groundwater flow and contaminant transport processes. Deep learning-based ensemble smoother methods have proven to be effective in estimating non-Gaussian K fields by assimilating hydrodynamic measurements. However, traditional hydrogeological investigations (e.g., hydraulic tomography) often suffer from the limited number of invasive borehole data (e.g., hydraulic head), making the non-Gaussian K inversion problem strongly underdetermined. As a non-invasive and high-sampling-density geophysical approach, the self-potential (SP) method can measure fluctuations of the electrical field caused by subsurface fluid movement. Thus, the recorded SP signals are electro-kinetic responses of the groundwater flow and can provide complementary information for delineating non-Gaussian heterogeneity. In this paper, we performed a joint hydrogeophysical inversion of hydraulic-head and SP data to characterize the non-Gaussian K field within a deep learning-based inversion framework. We conducted three synthetic cases in a 3-D non-Gaussian aquifer with 16,000 unknown K, to assess the ability of the proposed joint inversion framework for K imaging by assimilating different types of data. The results show that using limited hydraulic-head data alone may capture the large-scale features of the non-Gaussian reference field but might lose fine features. By integrating the hydraulic-head and SP data, the non-Gaussian K field can be reconstructed with an improved accuracy and reduced estimation uncertainty. The low-cost SP method opens up new opportunities for non-Gaussian K characterization and fluid-dynamic monitoring.

关键词： Joint hydrogeophysical inversion Non-Gaussian hydraulic conductivity Self-potential Hydraulic tomography convolutional variational autoencoder

来源：评论

学校读者我要写书评

暂无评论

Normal Appearance autoencoder for Lung Cancer Detection and Segmentation 22nd

Normal Appearance Autoencoder for Lung Cancer Detection and ...

引用

10th International Workshop on Machine Learning in Medical Imaging (MLMI) / 22nd International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI)

作者： Astaraki, Mehdi Toma-Dasu, Iuliana Smedby, Orjan Wang, Chunliang KTH Royal Inst Technol Dept Biomed Engn & Hlth Syst Halsovagen 11C S-14157 Huddinge Sweden Karolinska Univ Sjukhuset Karolinska Inst Dept Oncol Pathol S-17176 Solna Sweden

ISBN: (纸本)9783030322267;9783030322250

One of the major differences between medical doctor training and machine learning is that doctors are trained to recognize normal/healthy anatomy first. Knowing the healthy appearance of anatomy structures helps doctors to make better judgement when some abnormality shows up in an image. In this study, we propose a normal appearance autoencoder (NAA), that removes abnormalities from a diseased image. This autoencoder is semi-automatically trained using another partial convolutional in-paint network that is trained using healthy subjects only. The output of the autoencoder is then fed to a segmentation net in addition to the original input image, i.e. the latter gets both the diseased image and a simulated healthy image where the lesion is artificially removed. By getting access to knowledge of how the abnormal region is supposed to look, we hypothesized that the segmentation network could perform better than just being shown the original slice. We tested the proposed network on the LIDC-IDRI dataset for lung cancer detection and segmentation. The preliminary results show the NAA approach improved segmentation accuracy substantially in comparison with the conventional U-Net architecture.

关键词： Lung nodule segmentation Anomaly detection convolutional variational autoencoder

来源：评论

学校读者我要写书评

暂无评论

DEEP variational FILTER LEARNING MODELS FOR SPEECH RECOGNITION 44

DEEP VARIATIONAL FILTER LEARNING MODELS FOR SPEECH RECOGNITI...

引用

44th IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

作者： Agrawal, Purvi Ganapathy, Sriram Indian Inst Sci Elect Engn LEAP Lab Bangalore Karnataka India

ISBN: (纸本)9781479981311

We present a novel approach to derive robust speech representations for automatic speech recognition (ASR) systems. The proposed method uses an unsupervised data-driven modulation filter learning approach that preserves the key modulations of speech signal in spectro-temporal domain. This is achieved by a deep generative modeling framework to learn modulation filters using convolutional variational autoencoder (CVAE). A skip connection based CVAE enables the learning of multiple irredundant modulation filters in the time and frequency modulation domain using temporal and spectral trajectories of input spectrograms. The learnt filters are used to process the spectrogram features for ASR training The ASR experiments are performed on Aurora-4 (additive noise with channel artifact) and CHiME-3 (additive noise with reverberation) databases. The results show significant improvements for the proposed CVAE model over the baseline features as well as other robust front-ends (average relative improvements of 9% in word error rate over base-line features on Aurora-4 database and 23% on CHiME-3 database). In addition, the performance of the proposed features is highly beneficial for semi-supervised training of ASR when reduced amounts of labeled training data are available (average relative improvements of 29% over baseline features on Aurora-4 database with 30% of the labeled training data).

关键词： Unsupervised filter learning convolutional variational autoencoder skip connections modulation filtering robust speech recognition

来源：评论

学校读者我要写书评

暂无评论

Unsupervised Raw Waveform Representation Learning for ASR 20

Unsupervised Raw Waveform Representation Learning for ASR

引用

Interspeech Conference

作者： Agrawal, Purvi Ganapathy, Sriram Indian Inst Sci Learning & Extract Acoust Patterns LEAP Lab Dept Elect Engn Bengaluru 560012 India

In this paper, we propose a deep representation learning approach using the raw speech waveform in an unsupervised learning paradigm. The first layer of the proposed deep model performs acoustic filtering while the subsequent layer performs modulation filtering. The acoustic filterbank is implemented using cosine-modulated Gaussian filters whose parameters are learned. The modulation filtering is performed on log transformed outputs of the first layer and this is achieved using a skip connection based architecture. The outputs from this two layer filtering are fed to the variational autoencoder model. All the model parameters including the filtering layers are learned using the VAE cost function. We employ the learned representations (second layer outputs) in a speech recognition task. Experiments are conducted on Aurora-4 (additive noise with channel artifact) and CHiME-3 (additive noise with reverberation) databases. In these experiments, the learned representations from the proposed framework provide significant improvements in ASR results over the baseline filterbank features and other robust front-ends (average relative improvements of 16% and 6% in word error rate over baseline features on clean and multi-condition training, respectively on Aurora-4 dataset, and 21% over the baseline features on CHiME-3 database).

关键词： Unsupervised representation learning raw speech waveform convolutional variational autoencoder cosine-modulated Gaussian filterbank speech recognition

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：