检索结果-内蒙古大学图书馆

Neural-Network-Based Direction-of-Arrival Estimation for Reverberant Speech - The Importance of Energetic, Temporal, and Spatial Information

引用

IEEE-ACM TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE processing 2024年 32卷 1298-1309页

作者： Ben Zaken, Orel Kumar, Anurag Tourbabin, Vladimir Rafaely, Boaz Ben Gurion Univ Negev Dept Elect & Comp Engn IL-8410501 Beer Sheva Israel Facebook Inc Facebook Real Labs Redmond WA 20004 USA Facebook Seattle Seattle WA 98101 USA

Direction-of-arrival (DOA) estimation is a fundamental task in audio signal processing that becomes difficult in real-world environments due to the presence of reverberation. To address this difficulty, Direct-Path Dominance (DPD) tests have been proposed as an effective approach for detecting time-frequency (TF) bins dominated by direct sound, which contain accurate DOA information. These have been found to be particularly efficient when working with spherical arrays. While methods based on neural networks (NNs) have been developed to estimate the DOA, they have limitations such as the need for a large training database, and often understanding of the system's operation is lacking. This work proposes two novel DPD-test methods based on a model-based deep learning approach that combines the original DPD-test model with a data-driven system. Thus, it is possible to preserve the robustness of the original DPD-test across acoustic environments, while using a data-driven approach to better extract useful information about the direct sound, thereby enhancing the original method's performance. In particular, the paper investigates how energetic, temporal and spatial information contribute to the identification of TF-bins dominated by the direct signal. The proposed methods are trained on simulated data of a single sound source in a room, and evaluated on simulated and real data. The results show that energetic and temporal information provide new information about direct sound, which has not been considered in previous works and can improve its performance.

关键词： Direction-of-arrival estimation Estimation Training Arrays time-frequency analysis Reverberation deep learning Speaker localization direction-of-arrival (DOA) spherical arrays machine learning deep learning neural network (NN) multilayer perceptron (MLP) long short-term memory (LSTM)

来源：评论

学校读者我要写书评

暂无评论

DEPICTER: deep representation clustering for histology annotation

引用

COMPUTERS IN BIOLOGY AND MEDICINE 2024年 170卷 108026-108026页

作者： Chelebian, Eduard Avenel, Chirstophe Ciompi, Francesco Waehlby, Carolina Uppsala Univ Dept Informat Technol Uppsala Sweden Radboud Univ Nijmegen Dept Pathol Nijmegen Med Ctr Nijmegen Netherlands

Automatic segmentation of histopathology whole -slide images (WSI) usually involves supervised training of deep learning models with pixel -level labels to classify each pixel of the WSI into tissue regions such as benign or cancerous. However, fully supervised segmentation requires large-scale data manually annotated by experts, which can be expensive and time-consuming to obtain. Non -fully supervised methods, ranging from semi -supervised to unsupervised, have been proposed to address this issue and have been successful in WSI segmentation tasks. But these methods have mainly been focused on technical advancements in algorithmic performance rather than on the development of practical tools that could be used by pathologists or researchers in real -world scenarios. In contrast, we present DEPICTER (deep rEPresentatIon ClusTERing), an interactive segmentation tool for histopathology annotation that produces a patch -wise dense segmentation map at WSI level. The interactive nature of DEPICTER leverages self- and semi -supervised learning approaches to allow the user to participate in the segmentation producing reliable results while reducing the workload. DEPICTER consists of three steps: first, a pretrained model is used to compute embeddings from image patches. Next, the user selects a number of benign and cancerous patches from the multi -resolution image. Finally, guided by the deep representations, label propagation is achieved using our novel seeded iterative clustering method or by directly interacting with the embedding space via feature space gating. We report both real-time interaction results with three pathologists and evaluate the performance on three public cancer classification dataset benchmarks through simulations. The code and demos of DEPICTER are publicly available at https://***/eduardchelebian/depicter.

关键词： Interactive annotation Histology Self-supervised learning Clustering

来源：评论

学校读者我要写书评

暂无评论

A deep learning framework for suppressing prestack seismic random noise without noise-free labels

引用

Energy Geoscience 2024年第3期5卷 261-274页

作者： Han Wang Jie Zhang Petroleum Exploration and Production Research Institute SINOPECBeijing100083China Geophysical Research Institute School of Earth and Space SciencesUniversity of Science and Technology of ChinaHefeiAnhui230026China

Random noise attenuation is significant in seismic data *** deep learning-based denoising methods have been widely developed and applied in recent *** practice,it is often time-consuming and laborious to obtain noise-free data for supervised ***,we propose a novel deep learning framework to denoise prestack seismic data without clean labels,which trains a high-resolution residual neural network(SRResnet)with noisy data for input and the same valid data with different noise for *** valid signals in noisy sample pairs are spatially correlated and random noise is spatially independent and unpredictable,the model can learn the features of valid data while suppressing random *** data targets are generated by a simple conventional method without fine-tuning *** initial estimates allow signal or noise leakage as the network does not require clean *** Monte Carlo strategy is applied to select training patches for increasing valid patches and expanding training *** learning is used to improve the generalization of real data *** synthetic and real data tests perform better than the commonly used state-of-the-art denoising methods.

关键词： Data processing Denoising Signal processing Seismics deep learning

来源：评论

学校读者我要写书评

暂无评论

Enhancing Process image Analysis with Adaptive Thresholding and deep learning Techniques

Enhancing Process Image Analysis with Adaptive Thresholding ...

引用

2024 IEEE International Conference on Imaging Systems and Techniques, IST 2024

作者： Park, Hyeong Hu Hwang, Sung Jae Kim, Do Gyun Secondary Cell Equipment Manufacturing Division Software Team APSystems Gyeonggi-do Korea Republic of

ISBN: (纸本)9798350378214

This paper presents a deep learning model specifically designed to effectively classify display Mura images. The model leverages advanced deep learning techniques and computer vision methods to identify and categorize various types of Mura based on their unique digital signatures and visual patterns. It aims to provide fast and accurate classification results, enabling real-time processing of large-scale image data. The model is expected to significantly enhance content management and user experience with display Mura, and it can be innovatively applied across various fields of image classification. © 2024 IEEE.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

Video Analysis Method of Basketball Training Based on deep learning in the Context of Online Gaming

引用

Computer-Aided Design and Applications 2024年第s5期21卷 83-90页

作者： Qiushi, Li Henan Polytechnic Institute NanYang China

In order to tackle some issues of the inadequate data clustering in the original basketball shooting track capture and counter-capture method, a novel approach is proposed. This method utilizes the background differentiation technique for automated capture of basketball shooting trajectories. A sensor network is designed to capture the trajectory, ensuring improved accuracy and data integrity. The design process emphasizes the careful placement of sensors and the post-acquisition data processing. real-time and effective data acquisition is achieved through the use of inertial sensors. The trapper module handles data processing and trajectory capture. By employing the background differentiation technique, the target trajectory is detected and processed, enabling automatic capture of the shooting trajectory. This method significantly enhances track capture accuracy. Upon widespread adoption, it holds the potential to greatly improve the track capture impact in the field of basketball. © 2024 CAD Solutions,.

关键词： deep learning

来源：评论

学校读者我要写书评

暂无评论

Research on Water Gauge Recognition Based on deep learning and image processing 2021

Research on Water Gauge Recognition Based on Deep Learning a...

引用

2021 4th International Conference on E-Business, Information Management and Computer Science

作者： Yuehao Tang Chong Guo Changjiang Institute of Technology China Wuhan Business University China

ISBN: (纸本)9781450395687

Abstract: In order to improve the intellective level of water resources management, a real-time water level recognition method based on deep-learning algorithms and image-processing techniques is proposed in this paper. The recognition process is composed of four steps. Firstly, for the purpose of digit detection, YOLO-v3 model is deployed for extracting numbers from the water gauges. Then, the cropped number images are fed into the LSTM + CTC model as training samples so that digits can be recognized. In the third step, Hough transform are adopted to correct the tilt of water gauge in terms of the vertical edge feature. Morphological operation, associated with horizontal projection would position upper and lower edge of water gauge to recognize the scale lines correctly. Water level could be determined correspondingly. Model application shows that the recognition model has satisfying accuracy and efficiency, with potential being applied in practice.

关键词： Water Gauge image processing Number Recognition deep learning Water Level Number Detection

来源：评论

学校读者我要写书评

暂无评论

A Fast Unsupervised image Stitching Model Based on Homography Estimation

引用

IEEE SENSORS JOURNAL 2024年第18期24卷 29452-29467页

作者： Ni, Jianjun Li, Yingqi Ke, Chunyan Zhang, Ziru Cao, Weidong Yang, Simon X. Hohai Univ Coll Artificial Intelligence & Automat Changzhou 213200 Jiangsu Peoples R China Univ Guelph Sch Engn Adv Robot & Intelligent Syst ARIS Lab Guelph ON N1G 2W1 Canada

image stitching is the synthesis of multiple partial image segments into a complete and continuous panoramic image through effective image alignment and seamless fusion techniques. It can achieve a wider field of view and richer information for display and analysis. Most deep learning-based image stitching methods have significant advantages in improving accuracy, but they are not suitable for real-time applications due to multiple iterations of computation or deeper network depth. To deal with this problem, a fast unsupervised image stitching model is proposed in this article. In the proposed model, an adaptive feature extraction module (FEM) for deformation is designed, and then a fast unsupervised learning-based image alignment network is proposed. In addition, a stitching restoration network with a smaller number of parameters is presented to remove the redundant and unnecessary sampling and convolution operations in general deep learning-based models. Finally, some experiments are conducted on both the synthetic and real-scene datasets. The total stitching accuracy of the proposed model is higher, and the details of the output images are clearer. The proposed can achieve 1.79, 26.54, and 0.86 in RMSE, peak signal-to-noise ratio (PSNR), and structural similarity (SSIM) on the alignment results, respectively, which are better than those of the state-of-the-art methods. Furthermore, the comparison results prove that the proposed model can effectively reduce memory loss, and achieve a fast unsupervised image stitching, with a very small model size.

关键词： Deformation invariance fast alignment homography estimation image stitching

来源：评论

学校读者我要写书评

暂无评论

A novel frequency-division deep-learning approach for magnetotelluric data quality enhancement

引用

GEOPHYSICS 2025年第3期90卷 WA169-WA187页

作者： Yu, Nian Ji, Mingjie Zhang, Chao Ye, Yi Zhou, Wei Chongqing Univ Sch Elect Engn Chongqing Peoples R China State Key Lab Coal Mine Disaster Dynam & Control Chongqing Peoples R China Chongqing Univ Sch Resources & Safety Engn Chongqing Peoples R China

High signal-to-noise ratio magnetotelluric (MT) data are crucial for accurately interpreting subsurface structures. Recently, deep learning has become popular for MT denoising due to its ability to avoid parameter tuning and enable real-time processing. These methods typically fit or predict signals in noisy segments after identifying and segmenting signal and noise in the time domain. However, these methods struggle to preserve low-and high-frequency signals effectively due to high noise levels in these segments. To address this issue, we develop a novel deep-learning denoising method that separately recovers low-and high-frequency signals using distinct strategies. Low-frequency signals are fitted using an inverse autoencoder with a channel attention mechanism, effectively removing high-frequency components. High-frequency signals are then predicted using a bidirectional long short-term memory network combined with a squeeze-and-excitation mechanism, enhancing prediction by considering global and local signal characteristics. In addition, we introduce the multivariate state estimation technique (MSET) for real-time signal-noise identification. MSET analyzes residuals after separating low-frequency signals to identify noise. Denoising is performed only on segments with significant noise, preserving more effective signals. Finally, the fitted low-frequency dominant and predicted high-frequency components are combined to form the denoised MT signals. This combined approach significantly improves the restoration quality of effective signals compared with existing methods. Experimental results demonstrate that our method exhibits superior denoising capabilities in quantitative and qualitative evaluations, including apparent resistivity-phase curves and polarization direction analysis, offering enhanced performance over current deep-learning methods.

关键词： magnetotelluric

来源：评论

学校读者我要写书评

暂无评论

real-time Intrusion Detection in Power Grids Using deep learning: Ensuring DPU Data Security

HighTech and Innovation Journal

引用

HighTech and Innovation Journal 2024年第3期5卷 814-827页

作者： Xiao, Maoran Zhou, Qi Zhang, Zhen Yin, Junjie State Grid Jiangsu Electric Power Co Ltd. Limited Information and Telecommunication Branch Jiangsu Nanjing210000 China State Grid Jiangsu Electric Power Co Ltd. Wuxi Power Supply Branch Jiangsu Wuxi214000 China

deep learning technologies have revolutionized the management of energy, energy consumption, and data security within smart grids through non-intrusive load monitoring (NILM). This paper explores the use of deep learning for real-time intrusion detection in power grids with a primary focus on safeguarding the integrity and security of Data processing Units (DPUs). An evaluation of various machine learning models, including Support Vector Machine (SVM), Linear Discriminant Analysis (LDA), Decision Trees, and Random Forests, is conducted to detect various types of intrusions, including Fault, Injection, Masquerade, Normal, and Replay. Random Forest produced AUC values of 1.00 for all classes and an overall F1-score of 0.99 for all classes. The Decision Tree model also shows robust performance for detecting Fault and Injection intrusions (AUC = 0.98), with an overall F1-score of 0.94. However, the LDA and SVM models do not perform well in detecting Injection intrusions with overall F1-scores of 0.83 and 0.86. Advances in machine learning can be used to improve smart grid security, reliability, and efficiency, according to this study. These findings highlight the potential of advanced machine learning techniques to enhance smart grid reliability and efficiency. © 2024, Ital Publication. All rights reserved.

关键词： Data integrity

来源：评论

学校读者我要写书评

暂无评论

Enhanced SCNN-Based Hybrid Spatial-Temporal Lane Detection Model for Intelligent Transportation Systems

引用

IEEE ACCESS 2024年 12卷 40075-40091页

作者： Li, Jingang Ma, Chenxu Han, Yonghua Mu, Haibo Jiang, Lurong Zhejiang Sci Tech Univ Sch Informat Sci & Engn Hangzhou 310018 Peoples R China Hangzhou Hikvis Digital Technol Co Ltd Hangzhou 310018 Peoples R China

Accurate and timely lane detection is imperative for the seamless operation of autonomous driving systems. In this study, leveraging the gradual variation of lane features within a defined range of width and length, we introduce an enhanced Spatial-Temporal Recurrent Neural Network (SCNN) framework. This framework serves as the cornerstone of an innovative hybrid spatial-temporal model for lane detection, which is tailored to address the prevalent issues of substandard detection performance and insufficient real-time processing in intricate scenarios, such as those involving lane erosion and inconsistent lighting conditions, which often challenge conventional models. With the foundational understanding that lanes manifest as continuous lines, we employ a temporal sequence of lane imagery as the input to our model, thereby ensuring a rich provision of feature information. The model adopts an encoder-decoder structure and integrates a Spatial-Temporal Recurrent Neural Network module for the extraction of interrelated information from the image sequence. The model culminates in the output of the lane detection results for the terminal frame. The proposed lane detection model exhibits a commendable synthesis of accuracy and real-time efficiency, attaining an Accuracy of 97.87%, an F-1 -score of 0.943, and a FPS of 19.342 on the tvtLANE dataset and an Accuracy of 98.21%, an F-1 -score of 0.957 on the Tusimple dataset. These metrics signify a superior performance over a majority of the current lane detection methods.

关键词： Computer vision deep learning Lane detection Autonomous driving Hybrid power systems Spatiotemporal phenomena Intelligent transportation systems Recurrent neural networks deep learning lane detection

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：