检索结果-内蒙古大学图书馆

IEEE International Symposium on Parallel and Distributed processing with applications and IEEE International conference on Ubiquitous Computing and Communications (ISPA/IUCC)

作者： Yuan Liu Rong Xiang Qieshi Zhang Ziliang Ren Jun Cheng Guilin University of Electronic Technology Guilin China Key Laboratory of Human-Machine Intelligence-Synergy Shenzhen China

ISBN: (纸本)9781728152103

Loop closure detection plays a vital role in the visual Simultaneous Localization and Mapping (SLAM) system. Traditional Bag-of-Visual-Word (BoVW) methods for loop closure detection has been successfully applied to various SLAM systems. The Convolutional neural networks (ConvNets) based methods can automatically learn feature representation from original image, and it is more robust to illumination changes. However, these two methods all have some limitations. In order to overcome the shortcomings of the traditional methods extract insufficient features, and the ConvNets based methods unable to adapt complex environment changes, in this paper, we propose an improved hybrid deep learning architecture (HDLA) to generate high-level semantic image features for loop closure detection specifically. This network is based on hybrid ConvNet modified by us to especially cope with robust and real-time feature extraction. We present extensive experiments on three real world datasets to evaluate each of the specific challenges in loop closure detection. The results demonstrate that the proposed method achieves superior performances and provides a feasible solution for loop closure detection.

关键词： convolutional neural nets feature extraction image classification image representation learning (artificial intelligence) object detection robot vision SLAM (robots) robot vision Feature extraction image representation Object detection image classification surgical closure Loops Contraceptive Coil Feasible Solution original image simultaneous localization and mapping

来源：评论

学校读者我要写书评

暂无评论

Mobile Medical Assistant and Analytical System for Dengue Patients

Mobile Medical Assistant and Analytical System for Dengue Pa...

引用

Advancements in Computing (ICAC), 2019 International conference on

作者： K. T. K. Jayampathi M. A. C. Jananjaya E. P. C. Fernando Y. A. Liyanage M. G. N. M. Pemadasa G. W. D. A. Gunarathne Faculty of Computing Sri Lanka Institute of Information Technology Malabe Sri Lanka

ISBN: (纸本)9781665408639

Dengue fever is a vector-borne viral disease spread by the mosquito Aedes Aegypti. It is a public health problem, with an estimated 50-500 million infections each year and no effective vaccination. People's hectic schedules may not have enough time to see a doctor every time they have a fever. They may overlook their disease, believing it to be a common ailment. Prior medical assistance for dengue patients with fever to check their conditions reliably is a major problem. There is no easily accessible proper system to identify dengue patients at an early stage. This paper presents a mobile medical assistant and analytical system for dengue patients. With a novel approach, using the most appropriate technologies, the mobile application supports identifying dengue patients using the chatbot, analyzing skin conditions, analyzing blood reports, and analyzing dengue-infected areas' functionalities. The registered users can log in to the system and check their dengue condition. The development is carried out with Natural Language processing, artificial neural Network (ANN), Machine Learning, image processing, Convolutional neural Network (CNN), and Android technologies. A mobile application prototype is created and tested, with the possibility of future testing and implementation. The results show effective performances in analyzing dengue conditions.

关键词： Schedules Skin Mobile applications Vaccines Convolutional neural networks Reliability Viruses (medical)

来源：评论

学校读者我要写书评

暂无评论

Comprehensive investigation on Deep learning models: applications, Advantages, and Challenges

Comprehensive investigation on Deep learning models: Applica...

引用

International conference on Computing and Networking Technology (ICCNT)

作者： Rajiv Avacharmal Saigurudatta Pamulaparthyvenkata Dileep Kumar Pandiya R Krishna Prakash Piyush Ranjan Anandaganesh Balakrishnan S Maheswaran AI/ML expert University of Connecticut USA Independent Researcher ZoomInfo Technologies Department of Information Technology Sona College of Technology Salem Tamil Nadu IEEE USA Data Engineering American Water Department of Electronics and Communication Engineering Kongu Engineering College Perundurai Erode TamilNadu

ISBN: (数字)9798350370249

ISBN: (纸本)9798350370270

artificial intelligence (AI) and machine learning (ML) have been completely transformed by deep learning (DL), which provides unmatched power in handling massive, unstructured information from a variety of fields. CNNs, neural networks with recurrent connections (RNNs), generative models, (DRL), and deep learning via transfer are the main deep learning models that are covered in detail in this article. The structure, uses, advantages, and drawbacks of each model are thoroughly investigated. In this work, we use the Fruit-360 dataset, a benchmark picture classification dataset, for an empirical investigation. Six well-known deep learning architectures are compared and evaluated: Bidirectional LSTM, CNN, Simple RNN, (LSTM), (GRU), and Bidirectional GRU. Analysis is done on performance parameters including accuracy, precision, recall, and computing efficiency to have a better understanding of how well they work on various tasks. The purpose of this survey is to give scholars, professionals, and enthusiasts a thorough grasp of the various uses and capacities of deep learning models. The study provides insights that help progress the discipline and direct the selection of suitable models for certain real-world problems.

关键词： Deep learning Training Surveys Computational modeling image processing Focusing Data models Spatial databases Classification algorithms Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Production of MusicXML from Locally Inclined Sheet-music Photo image by Using Measure-based Multi-modal Deep-learning-driven Assembly Method

引用

Transactions of the Japanese Society for artificial Intelligence 2023年第3期38卷

作者： Shishido, Tomoyuki Fati, Fehmiju Tokushige, Daisuke Ono, Yasuhiro Kumazawa, Itsuo Biomedical AI Research Unit Tokyo Institute of Technology Shishido&Associates Japan SK Intellectual Property Law Firm Japan Enspirea LLC Japan Laboratory for Future Interdisciplinary Research of Science and Technology Tokyo Institute of Technology Japan

Deep learning has been applied to optical music sheet recognition (OMR). However, OMR processing from various sheet-music images still lacks precision to be widely applicable. We propose a measure-based multimodal deep-learning-driven assembly (MMdA) method enabling end-to-end OMR processing from various images including inclined photo images. Using this method, measures are extracted using a deep-learning model, aligned, and resized to be used for inference of given musical-symbol components by using multiple deep-learning models in sequence or in parallel. The use of each standardized measure enables efficient training of the deep-learning models and accurate adjustment of five staff lines in each measure, which enables locally inclined sheet-music images to be precisely positioned. Thus, a score can be reproduced from the inclined image with the proposed MMdA method while current OMR applications cannot. Multiple musical-symbol-component deep-learning feature-category models with a small number of feature types can represent a diverse set of notes and other musical symbols including chords. The proposed MMdA method provides a solution to end-to-end OMR processing and enhances the utility of OMR of mobile phone-based sheet-music photo images. © 2023, Japanese Society for artificial Intelligence. All rights reserved.

关键词： Music

来源：评论

学校读者我要写书评

暂无评论

Automated Object Detection and Recognition in Satellite imagery Using CNNs

Automated Object Detection and Recognition in Satellite Imag...

引用

Innovations in Communications, Electrical and Computer Engineering (ICICEC), International conference on

作者： Shanthi H J G. Shobana E. Kamalanaban J. Senthil Murugan S. Nandhinieswari R. Stalinbabu Department of Computer Applications Hindustan Institute of Technology and Science Deemed to be University Padur School of Computer Science Takshashila University Department of Computer Science and Engineering Vel Tech High Tech Dr.Rangarajan Dr.Sakunthala Engineering College Vel Tech High Tech Dr. Rangarajan Dr.Sakunthala Engineering College Kongunadu Arts and Science College Coimbatore Gounder Mills Coimbatore Artificial Intelligence and Machine Learning M. Kumarasamy College of Engineering Karur

ISBN: (数字)9798350376517

ISBN: (纸本)9798350376524

As increasingly high-resolution satellite photography becomes available, it has led to a growing need for efficient and accurate methods to analyse large databases and extract valuable information. This research presents a comprehensive approach for automated item identification and detection in satellite data using Convolutional neural networks (CNNs). This research abstract outlines the critical steps involved in a suggested technique. These steps encompass data cleaning and organisation, model architecture development, training and evaluation, post- processing, deployment, and subsequent management. This study investigates the ability of CNNs to learn intricate spatial characteristics. The research focuses on carefully selecting and annotating training and validation datasets incorporating items of interest. This study uses data augmentation methods to enhance the generalisation capabilities of the model. This research abstract discusses the importance of the preprocessing stage in ensuring the consistent performance of a model by normalising the input data. This research abstract discusses the components of the CNN architecture, namely convolutional, pooling, and fully connected layers. The CNN architecture is specifically designed to capture hierarchical characteristics. This study investigates the utilisation of a cutting-edge optimiser in the training process to optimise a meticulously constructed loss function. This study examines the integration of the model into real-world applications, considering the trade-off between processing speed and resource constraints as crucial considerations for successful deployment. The findings suggest that this methodology holds promise as a valuable tool in various fields.

关键词： Training Technological innovation Accuracy Satellites image recognition Transfer learning Data augmentation Data models Satellite images Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Learning compact representations of neural networks using DiscriminAtive Masking (DAM) 21

Learning compact representations of neural networks using Di...

引用

Proceedings of the 35th International conference on neural Information processing Systems

作者： Jie Bu Arka Daw M. Maruf Anuj Karpatne Virginia Tech

ISBN: (纸本)9781713845393

A central goal in deep learning is to learn compact representations of features at every layer of a neural network, which is useful for both unsupervised representation learning and structured network pruning. While there is a growing body of work in structured pruning, current state-of-the-art methods suffer from two key limitations: (i) instability during training, and (ii) need for an additional step of fine-tuning, which is resource-intensive. At the core of these limitations is the lack of a systematic approach that jointly prunes and refines weights during training in a single stage, and does not require any fine-tuning upon convergence to achieve state-of-the-art performance. We present a novel single-stage structured pruning method termed DiscriminAtive Masking (DAM). The key intuition behind DAM is to discriminatively prefer some of the neurons to be refined during the training process, while gradually masking out other neurons. We show that our proposed DAM approach has remarkably good performance over a diverse range of applications in representation learning and structured pruning, including dimensionality reduction, recommendation system, graph representation learning, and structured pruning for image classification. We also theoretically show that the learning objective of DAM is directly related to minimizing the L0 norm of the masking layer.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Efficient Bayesian learning curve extrapolation using prior-data fitted networks 23

Efficient Bayesian learning curve extrapolation using prior-...

引用

Proceedings of the 37th International conference on neural Information processing Systems

作者： Steven Adriaensen Herilalaina Rakotoarison Samuel Müller Frank Hutter Machine Learning Lab University of Freiburg

Learning curve extrapolation aims to predict model performance in later epochs of training, based on the performance in earlier epochs. In this work, we argue that, while the inherent uncertainty in the extrapolation of learning curves warrants a Bayesian approach, existing methods are (i) overly restrictive, and/or (ii) computationally expensive. We describe the first application of prior-data fitted neural networks (PFNs) in this context. A PFN is a transformer, pre-trained on data generated from a prior, to perform approximate Bayesian inference in a single forward pass. We propose LC-PFN, a PFN trained to extrapolate artificial right-censored learning curves generated from a parametric prior proposed in prior art using MCMC. We demonstrate that LC-PFN can approximate the posterior predictive distribution over learning curves more accurately than MCMC, while being over 10000 times faster. We also show that the same LC-PFN achieves competitive performance extrapolating a total of 20 000 real learning curves from four learning curve benchmarks (LCBench, NAS-Bench-201, Taskset, and PD1) that stem from training a wide range of model architectures (MLPs, CNNs, RNNs, and Transformers) on 53 different datasets with varying input modalities (tabular, image, text, and protein data). Finally, we investigate its potential in the context of model selection and find that a simple LC-PFN based predictive early stopping criterion obtains 2 - 6 × speed-ups on 45 of these datasets, at virtually no overhead.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Automated Classification of Test Tubes Based on Uncontrolled image Analysis

Automated Classification of Test Tubes Based on Uncontrolled...

引用

Metrology for Extended Reality, artificial Intelligence and neural Engineering (MetroXRAINE), IEEE International conference on

作者： Riccardo Balia Silvio Barra Alessandro Sebastian Podda Livio Pompianu Mara Sangiovanni Gianni Fenu Dept. of Mathematics and Computer Science (DMI) University of Cagliari Italy Dept. of Information Technology and Electrical Engin. (ITEE) University of Naples ‘Federico II’ Italy

ISBN: (数字)9798350378009

ISBN: (纸本)9798350378016

Metrology, the science of measurement, involves defining and establishing standards for quantifying physical quantities. A critical aspect of this discipline is to accurately measure and understand the relationships between object dimensions. However, determining dimensions from images poses several challenges, including optical distortions, lighting errors, and perspective issues, which can distort the perceived dimensions of objects. Additionally, proper calibration of images and knowledge of exact real-world reference dimensions are essential for achieving precise and accurate measurements. In this paper, we propose a solution, based on a Convolutional neural networks (CNNs), for classifying three different types of test tubes from images. The test tubes exhibit slight dimensional variations ( $\mathbf{16}\mathbf{mm}\times \mathbf{100}\mathbf{mm}, \mathbf{13}\mathbf{mm}\times \mathbf{100}\mathbf{mm}$ , and $\mathbf{13}\mathbf{mm}\times \mathbf{75}\mathbf{mm}$ ). The images are captured in uncontrolled conditions, without specific guidelines for the operator, and lack reference dimensions to aid in identifying the test tube type. To address the absence of reference dimensions, our approach relies on capturing and analyzing subtle visual cues and patterns unique to each test tube type. By training on a diverse set of samples, our model can generalize effectively to unseen images, providing robust and accurate classification results. Experimental evaluations demonstrate the effectiveness of our proposed method, achieving high classification accuracy even under challenging conditions with varying lighting, backgrounds, and orientations. The results underscore the potential of our approach in practical applications where manual classification based on visual inspection alone is unreliable or time-consuming.

关键词： Training Visualization Accuracy Lighting Optical distortion Metrology Predictive models Distortion measurement Standards Guidelines

来源：评论

学校读者我要写书评

暂无评论

"Refining image Quality: Deep neural Network applications in Plane Wave Medical Imaging"

"Refining Image Quality: Deep Neural Network Applications in...

引用

International conference on Circuit, Power and Computing Technologies (ICCPCT)

作者： D. Ferlin Deva Shahila L. Padma Suresh A. Merry Ida D. Lilly Evangelin P Aruna Jeyanthy Valantina Stephen Department of ECE Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology Chennai India Department of CSE Baselios Mathews II College of Engineering Kollam Chennai India Department of CSE Loyola Institute of Technology and Science Loyola Nagar Thovalai Kanya Kumari India Department of CSE Loyola Institute of Technology and Science Loyola Nagar Thovalai Chennai India Department of EEE Loyola Institute of Technology and Science Loyola Nagar Thovalai Kanya Kumari India Department of EEE Kalasalingam Academy of Research and Education Srivilliputhur Tamil Nadu India Department of EEE Saveetha School of Engineering Thandalam Chennai India

ISBN: (数字)9798350372816

ISBN: (纸本)9798350372823

With the goal to attain a high frame rate, plane wave imaging (PWI), a common swift ultrasonography imaging in medicine method, uses a solitary plane wave emission with no focal point. However, as compared to the widely utilized targeted line scan mode, the imaging quality is significantly reduced. Imaging performance can be enhanced with conventional adaptive beamformers, albeit at the expense of more processing. In the present study, we suggest employing a deep neural network (DNN) to maintain a high frame rate despite improving PWI performance. Specifically, the key contribution of this technology is The network input is the PWI response from a single point target, and the narrower scan response from the same point appears as the desired output. An amalgam of phantom investigations, in vivo research, and computations are used to assess the effectiveness of the suggested approach. For comparison coherence factor (CF), a prior anticipated deep learning-based technique, the delay-and-sum (DAS), and the DAS with a customized scan are employed. The performance is measured using numerical metrics, such as the speckle the contrast ratio (CR), signal-to-noise ratio, and contrast-to-noise ratio (CNR) (SSNR). The outcomes show that the suggested strategy is capable of achieving better contrast and resolution. In particular, the suggested approach outperforms the DAS across the board. While the CF offers a greater CR, the suggested The SSNR and CNR of this strategy are significantly greater. In addition, the total performance is comparable with concentrated scan performance and superior to the conventional deep learning technique. Furthermore, the suggested technique ensures good temporal resolution with less additional computation needed comparison to the DAS. The aforementioned results confirm that the suggested strategy is capable of producing high-quality images while preserving the high frame rate required for PWI.

关键词： In vivo image resolution Refining Phantoms artificial neural networks Ultrasonography Speckle

来源：评论

学校读者我要写书评

暂无评论

Editorial for Special Issue: Remote Sensing Based Building Extraction ii

引用

REMOTE SENSING 2023年第4期15卷 998-998页

作者： Tian, Jiaojiao Yan, Qin Awrangjeb, Mohammad Kallfelz (Sirmacek), Beril Demir, Nusret German Aerosp Ctr DLR Remote Sensing Technol Inst D-82234 Wessling Germany Chinese Acad Surveying & Mapping Beijing 100830 Peoples R China Griffith Univ Inst Integrated & Intelligent Syst Nathan Qld 4111 Australia Akdeniz Univ Dept Space Sci & Technol Remote Sensing Div TR-07058 Antalya Turkiye

Accurate building extraction from remotely sensed images is essential for topographic mapping, urban planning, disaster management, navigation, and many other applications [1]. The easily available very-high resolution 2D/3D dataset and the rapid development of image processing techniques, especially the convolutional neural networks (CNN) and deep learning techniques have further boosted the research on building-extraction-related topics. Especially in recent years, many research institutes and associations have provided open-source datasets and annotated training data to meet the demand for advanced artificial intelligence models, which brings new opportunities to develop advanced approaches for building extraction and monitoring. Hence, there are higher expectations of the efficiency, accuracy, and robustness of building extraction approaches. They should also meet the demand of processing large datasets at the city, national, and global levels. Moreover, challenges remain on transform learning and dealing with imperfect training data, as well as unexpected objects in urban scenes such as trees, clouds, and *** learning (DL) shows remarkable performance in extracting buildings from high-resolution remote sensing images. How to improve the performance of DL methods, especially the perception of spatial information, is worth further study. Paper [2] proposed a building extraction network (B-FGC-Net) with a feature highlighting, global awareness, and cross-level information fusion to achieve improved profitability of accurate extraction and information integration for both small- and large-scale buildings. Focusing on the promotion of the robustness of the interactive segmentation, Shu et al. [1] propose one Progress Guidance Representation Net (PGR-Net) to utilize the distance of newly added clicks to the boundary of the previous segmentation mask as an indication of the interactive segmentation progress, and this information is employed with the previous

关键词：

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：