检索结果-内蒙古大学图书馆

6th International conference on Advanced Informatics for Computing Research, ICAICR 2023

作者： verma, Divyanshu vijayarajan, v. Rajkumar, S. Ramesh Pathy, M. Shool of Computer Science and Engineering Vellore Institute of Technology Tamil Nadu Vellore632014 India School of Bio Sciences and Technology Vellore Institute of Technology Tamil Nadu Vellore632014 India

ISBN: (纸本)9783031840616

Hyperspectral imaging (HSI) emerges as a highly efficient means of information retention, particularly when leveraging the power of deep learning. The intricate task of non-destructively identifying rotten fruits is addressed through HSI, offering a non-invasive method for data collection that significantly streamlines the sorting process using advanced algorithms. Deep learning, employing neural network methodologies, facilitates the exploration of the intricate connection between spatial and spectral contexts within the data. The HSI system captures the wavelength information of refracted, reflected, and absorbed light across multiple bands. To address noise introduced by the machinery, a Standard Normal variate (SNv) filter is applied, resulting in a refined dataset represented as a hyperspectral cube (3D). The workflow involves the utilization of an image pre-training convolutional neural network (CNN) on the collected dataset, subsequently fine-tuned to serve as a robust classifier. In the proposed model, a substantial majority of the dataset (70%) is dedicated to training the neural network, with the remaining portion (30%) reserved for the validation process. The findings demonstrate that this analysis approach, leveraging the CNN algorithm, yields remarkably accurate results (>90%) and proves to be highly effective in discerning fruit quality. The presented methodology showcases a promising synergy between hyperspectral imaging and deep learning, offering a reliable solution for non-destructive fruit quality assessment. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Optimizing Soccer Match Prediction with Deep Convolutional Neural Networks and Grey Wolf Optimization

Optimizing Soccer Match Prediction with Deep Convolutional N...

引用

Advances in Electrical, Computing, Communication and Sustainable Technologies (ICAECT), International conference on

作者： Tushar Dhar Shukla v.Kanpur Rani Indrajeet Kumar Omar Mohsen Hussein Shaik Balkhis Banu Bhavani P University Institute of Physical Education & Sports Chandigarh University Mohali India Department of EEE Hindustan Institute of Technology and Science Chennai India Department of CSE Graphic Era Hill University Dehradun Graphic Era Deemed to be University Dehradun Uttarakhand India Department of Biomedical Engineering Al-Farahidi University Baghdad Iraq Department of Physiotherapy Fatima College of Health Sciences Al Ain UAE Dept of AI & DS K. Ramakrishnan college of Engineering Trichy India

ISBN: (数字)9798350356236

ISBN: (纸本)9798350356243

Predicting soccer match outcomes has become an essential area of interest in sports analytics, yet many existing studies rely on traditional machine learning methods that may not adequately capture the complex dynamics inherent in the sport. This study aims to address these limitations by proposing a novel approach utilizing a Deep Convolutional Neural Network (DCNN) optimized through Grey Wolf Optimization (GWO) to significantly enhance prediction accuracy. The methodology leverages a comprehensive dataset from the European Soccer database, gathered from Kaggle, which includes detailed match statistics, player attributes, and betting odds, allowing for a richer analysis. The results of the study indicate that the proposed DCNN model outperforms several baseline models, achieving notable improvements in accuracy, precision, and F1-scores, thereby showcasing its ability to learn complex patterns and relationships within the data effectively. Despite these promising findings, the study acknowledges certain limitations, such as the potential for further enhancement through the integration of additional features like spatiotemporal tracking data and more detailed player performance metrics. The outcomes suggest that the combination of advanced deep learning techniques with effective optimization strategies can yield substantial improvements in predictive analytics in sports. In conclusion, this research contributes significantly to the field by demonstrating the potential of deep learning in the accurate prediction of soccer match outcomes. It encourages future exploration of diverse datasets and methodologies to further refine predictive capabilities, ultimately supporting more informed decision-making processes for teams and coaches within the competitive landscape of soccer. Overall, the study underscores the transformative potential of machine learning in sports analytics.

关键词： Deep learning Accuracy Europe Predictive models data models Spatiotemporal phenomena Convolutional neural networks Predictive analytics Optimization Sports

来源：评论

学校读者我要写书评

暂无评论

Explainable AI with SHAP and Grad-CAM: A Novel Approach to visualizing Cassava Leaf Health using CIvE, ECI Indices, and an Advanced UNet Framework

Explainable AI with SHAP and Grad-CAM: A Novel Approach to V...

引用

Innovation in Computing and Engineering (ICE), International conference on

作者： R Srivats Abhiram Sharma Kalyanasundaram v Srisakthi Saravanan School of Computer Science and Engineering Vellore Institute of Technology Chennai India

ISBN: (数字)9798350380460

ISBN: (纸本)9798350380477

This paper presents an in-depth study on the classification and segmentation of both healthy and diseased Cassava leaves by visualizing traditional image processing techniques with deep learning approaches, and introducing explainability to enhance model transparency. Agricultural monitoring typically relies on expensive multispectral sensors to assess plant health, but this research explores the use of affordable RGB cameras. Two vegetation indices, the Elliptical Color Index (ECI) and Color Index for vegetation Extraction (CIvE), are applied to extract vegetation features from RGB images, with Otsu thresholding used to create binary masks for segmentation. While CIvE consistently outperforms ECI in segmenting leaf regions, both methods struggle to distinguish between healthy and diseased leaves effectively. To address these limitations, a UNet deep learning model was trained on the binary masks generated by the indices, achieving significantly improved segmentation accuracy through techniques like data augmentation and optimized hyperparameters. UNet outperformed traditional methods, reaching 97.36% accuracy, 93.14% precision, 94.27% recall, an F1-score of 96.11%, an IoU of 91.08%, and a Dice coefficient of 92.32%. A novel aspect of this study is the integration and comparison of Explainable AI (XAI) techniques, including Grad-CAM and SHAP, to provide visual insights into how the UNet model distinguishes between healthy and diseased regions. These tools offer a deeper understanding of the model’s decision-making process, with SHAP proving particularly effective in highlighting critical areas of focus. This comprehensive evaluation of traditional image processing, deep learning, and XAI methods contributes to a more accurate and comprehensive solution for plant health monitoring, addressing both performance and transparency challenges in agricultural applications.

关键词： Deep learning Image segmentation Accuracy Image color analysis Plants (biology) vegetation mapping Feature extraction Indexes Monitoring Diseases

来源：评论

学校读者我要写书评

暂无评论

One-Shot Learning for Pose-Guided Person Image Synthesis in the Wild

One-Shot Learning for Pose-Guided Person Image Synthesis in ...

引用

International conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Dongqi Fan Tao Chen Mingjie Wang Rui Ma Qiang Tang Zili Yi Qian Wang Liang Chang University of Electronic Science and Technology of China China Mobile Zijin Innovation Institute Zhejiang Sci-Tech University Jilin University University of British Columbia NanJing University China Mobile Research Institute

ISBN: (数字)9798350368741

ISBN: (纸本)9798350368758

Current Pose-Guided Person Image Synthesis (PGPIS) methods depend heavily on large amounts of labeled triplet data to train the generator in a supervised manner. However, they often falter when applied to in-the-wild samples, primarily due to the distribution gap between the training datasets and real-world test samples. While some researchers aim to enhance model generalizability through sophisticated training procedures, advanced architectures, or by creating more diverse datasets, we adopt the test-time fine-tuning paradigm to customize a pre-trained Text2Image (T2I) model. However, naively applying test-time tuning results in inconsistencies in facial identities and appearance attributes. To address this, we introduce a visual Consistency Module (vCM), which enhances appearance consistency by combining the face, text, and image embedding. Our approach, named OnePoseTrans, requires only a single source image to generate high-quality pose transfer results, offering greater stability than state-of-the-art data-driven methods. For each test case, OnePoseTrans customizes a model in around 48 seconds with an NvIDIA v100 GPU.

关键词： Training visualization Image synthesis Graphics processing units One shot learning Signal processing Stability analysis Generators Speech processing Tuning

来源：评论

学校读者我要写书评

暂无评论

A New Hybrid Assistive Module Towards Text Recognition and Speech Conversion System from Real Time Acquired Images to visually Impaired Peoples

A New Hybrid Assistive Module Towards Text Recognition and S...

引用

Sentiment analysis and Deep Learning (ICSADL), International conference on

作者： B. Gobika S. Megha v. Sivasakthi K. Tharageswari Computer Science and Engineering Karpagam Academy of Higher Education Coimbatore India

ISBN: (数字)9798331523923

ISBN: (纸本)9798331523930

Identification and classification of objects containing text is an essential step to assist the visual impaired peoples towards analyzing the acquired image through camera. In existing, familiar object recognition approaches such as vGG-16, DenseNet-121, ResNet-18 and YOLO were employed to identify and classify the object present in the acquired images with optimal speed and greater accuracy. Further those model produces less mean squared error to the captured imaged by camera. Despite of several advantages, those models are time consuming and it still produces the moderate accuracy on the dense distributions of the image captured with multiple objects and different dimension of the text present in those object. In order to tackle those challenges, an a new hybrid assistive module is designed and represented as Text Inferred Yolo model as YOLOv10. YOLOv10 has been employed to detect and select the multiple objects containing text in the image to assist the blind people with accurate text recognition using Tesseract OCR engine. Initially camera captured image or gathered image from coco dataset is partitioned into residual blocks in grid format containing equal dimensions. Partitioned image cells will be employed to the bounding box approach to extract the features of the grid. The model identifies the text features of the object and will be represented in form of text feature map using the layers of convolution neural network. Further Linear Convolution Neural Network has been incorporated to produce the text labels for the feature map containing various sizes through its fully connected layer. Linear CNN also smoothens the text edges of the class labels on processing it through normalization layer to effectively categorize characters in the extracted regions of the image. Sigmoid activation unit has used to generate the class label for the max pooled data. Finally, the segmented text is exposed to the blind people with sound using speaker and google text to speech

关键词： YOLO Accuracy Text recognition Convolution Neural networks Text detection Blindness Feature extraction Cameras Object recognition

来源：评论

学校读者我要写书评

暂无评论

Association analysis for visual exploration of Multivariate Scientific data Sets

引用

IEEE TRANSACTIONS ON visualIZATION AND COMPUTER GRAPHICS 2016年第1期22卷 955-964页

作者： Liu, Xiaotong Shen, Han-Wei Ohio State Univ Columbus OH 43210 USA

The heterogeneity and complexity of multivariate characteristics poses a unique challenge to visual exploration of multivariate scientific data sets. as it requires investigating the usually hidden associations between different variables and specific scalar values to understand the data's multi-faceted properties. In this paper, we present a novel association analysis method that guides visual exploration of scalar-level associations in the multivariate context. We model the directional interactions between scalars of different variables as information flows based on association rules. We introduce the concepts of informativeness and uniqueness to describe how information flows between scalars of different variables and how they are associated with each other in the multivariate domain. Based on scalar-level associations represented by a probabilistic association graph, we propose the Multi-Scalar Informativeness-Uniqueness (MSIU) algorithm to evaluate the informativeness and uniqueness of scalars. We present an exploration framework with multiple interactive views to explore the scalars of interest with confident associations in the multivariate spatial domain, and provide guidelines for visual exploration using our framework. We demonstrate the effectiveness and usefulness of our approach through case studies using three representative multivariate scientific data sets.

关键词： Multivariate data association analysis visual exploration multiple views

来源：评论

学校读者我要写书评

暂无评论

vMambaDA: visual State Space Model with Unsupervised Domain Adaptation in Cervical Cancer

VMambaDA: Visual State Space Model with Unsupervised Domain ...

引用

International Image Processing, Applications and Systems conference (IPAS)

作者： Hoang-Nguyen vu Thi-Ngoc-Truc Nguyen Minh-Duc Bui Kim-Ngan Ngoc Huynh Thanh Minh Nguyen Hoang Tran Minh Duy Nguyen v. AI Lab - AI Vietnam Ho Chi Minh City Vietnam Ho Chi Minh City University of Science Ho Chi Minh City Vietnam Ho Chi Minh City Medicine and Pharmacy University Ho Chi Minh City Vietnam Trung Vuong Hospital Ho Chi Minh City Vietnam Industrial University of Ho Chi Minh City Ho Chi Minh City Vietnam

ISBN: (数字)9798331506520

ISBN: (纸本)9798331506537

Cervical cancer remains a major public health concern in vietnam, ranking second only to breast cancer among women. Early detection through screening, particularly HPv testing and Pap smears, is critical in reducing cervical cancer mortality. While deep learning has shown great promise in medical image analysis, particularly for detecting cervical cancer, challenges remain, especially when models are applied to real-world datasets with limited labeled data due to privacy concerns and expensive annotations. This research offers an effective cervical cancer diagnostic approach that combines Sliced Wasserstein Distance (SWD), Maximum Classifier Discrepancy (MCD), and the vMamba model. The vMambaDA model learns domain-invariant features and adjusts to domain shifts to address the problems of accuracy, generalization, and speed in cervical cancer screening. vMambaDA has proven to be more adept at handling the intricacies of medical images than earlier models by exhibiting better classification accuracy and sensitivity through extensive testing on both public and private datasets. This automated method could enhance the early detection of cervical cancer and intervention by pushing the limits of domain adaptation in cytopathology.

关键词： Deep learning Adaptation models visualization Accuracy Sensitivity Mortality Public healthcare Cervical cancer Medical diagnostic imaging Testing

来源：评论

学校读者我要写书评

暂无评论

Leveraging Instruction-based Learning for Effective UAv Traffic Management in Dynamic Environment

Leveraging Instruction-based Learning for Effective UAV Traf...

引用

Multi-Agent Systems for Collaborative Intelligence (ICMSCI), International conference on

作者： R.S. Ramya Keren Lois Daniel S. Rajkumar Sruthi S Madhavan Akshya. J Mani Deepak Choudhry Department of Computer Science and Engineering Sri Ramakrishna Engineering College Coimbatore Department of Computer Science and Business Systems KGiSL Institute of Technology Coimbatore Department of Information Technology Sona College of Technology Salem Department of Artificial Intelligence and Data Science Nehru Institute of Engineering and Technology Coimbatore Department of Computational Intelligence SRM Institute of Science and Technology Chennai Department of Computing Technologies SRM Institute of Science and Technology Chennai

ISBN: (数字)9798331509828

ISBN: (纸本)9798331509835

With the popularization of UAvs (Unmanned Aerial vehicles) in surveillance, delivery services, environmental monitoring, and increasingly other operational aspects, besides the need to take off and land, efficient traffic management becomes integral for assuring and securing further development and exploitation. Most of the existing traffic management systems for UAvs depend on traditional rule-based algorithms or simpler machine learning techniques, such as DQN or PPO. However, these systems also have enormous challenges: dynamic environments, high-density traffic, and robust decision-making over unpredictable weather and network conditions. In addition, the above traditional models usually experience difficulties in balancing exploration and exploitation, leading to suboptimal routing and enlarged collision risks in complicated urban airspaces. These challenges have driven us to develop a new proposal: a UA v traffic management framework based on Double Deep Q-Network (DDQN) combined with Instruction-Based Learning. In this approach, advanced learning for DDQN optimizes processes at decision-making points for UAvs. The model proposed herein, along with comprehensive simulations and quantitative analyses, demonstrated higher performance in terms of leadership and cumulative rewards while keeping lower training loss compared to the classical algorithms such as DQN, A3C, and PPO. At the same time, the DDQN model presented a better success rate (95%), as also reflected in the balanced precision-recall curve showing an area under the curve of 0.90, over others in the on-point prediction of UAvstate and safety margin of collision. Our experimental results demonstrate that the proposed DDQN-based framework significantly enhances UA v traffic management by optimizing energy consumption in complex airspaces and reducing the risk of collision.

关键词： Training Analytical models Adaptation models Statistical analysis Atmospheric modeling Computational modeling Decision making Predictive models Autonomous aerial vehicles Safety

来源：评论

学校读者我要写书评

暂无评论

Enhanced Residual U-Net with Attention for Optic Disc and Cup Segmentation in Fundus Images 1st

Enhanced Residual U-Net with Attention for Optic Disc and...

引用

1st International conference on Computation of Artificial Intelligence and Machine Learning, ICCAIML 2024

作者： Kamath, Ritesh v. Kedari, Bhavana Girisha, S. Savitha, G. Shreesha, S. Department of Data Science and Computer Applications Manipal Institute of Technology Manipal Academy of Higher Education Manipal India Department of Information and Communication Technology Manipal Institute of Technology Manipal Academy of Higher Education Manipal India

ISBN: (纸本)9783031714832

Glaucoma, a progressive visual disease, damages the optic nerve, potentially resulting in irreversible vision loss. Effective treatment plans and prompt diagnosis are essential for positive patient outcomes. The principal glaucoma-related indication typically revolves around an odd ratio between the diameters of the optic cup and disc. In this regard, fundus images are analyzed for diagnosis of glaucoma. However, image quality, similar characteristics across classes, class imbalance issues, uneven shapes of optic disc and curves, and other factors make correct segmentation of optic disc and curve difficult from fundus images. To address these issues, the work proposes an enhanced Residual U-Net architecture capable of semantically segmenting the optic disc and cup from retinal fundus images. The proposed model has an Attention Module that can effectively extract the necessary feature map for segmentation from the input images. This improves the model’s capacity to pick out important features and eliminate extraneous noise. Using a publicly available dataset, the study analyses the efficacy of the proposed model using both quantitative and qualitative measures. The effectiveness of the model is examined in relation to various loss functions. To emphasize the significance of the suggested model, a comparative analysis between the proposed and the traditional model is also carried out. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Patient treatment

来源：评论

学校读者我要写书评

暂无评论

Breaking the carbon bind: How digitalization and energy transformation reshape carbon dependency based on wavelet and machine learning approaches

引用

Environmental Development 2025年 55卷

作者： Yu, Yang Jian, Xin Won, DooHwan Jahanger, Atif International Business School Hainan University Hainan Haikou 570228 China Department of Economics Pusan National University Busan 46241 South Korea

As global efforts to combat climate change intensify, digitalization has emerged as a crucial driver in reducing carbon dependency, with energy transformation also playing a significant role. Within this purview, this paper delves into the interplay among digitalization, energy transformation, and carbon dependency, utilizing Chinese country-level data spanning from 2005 to 2021. Recognizing potential variations in emission reduction policies over time, we employ the wavelet spectrum, wavelet local multiple correlation, wavelet coherence and machine learning methods for a comprehensive exploration. The outcomes of the wavelet spectrum analysis offer a visual depiction of the variable dynamics over time, furnishing substantial underpinning for discerning their intricate behaviors. Simultaneously, the findings from the wavelet local multiple correlation and wavelet coherence analyzes underscore disparities in the impacts of digitalization and energy transformation on carbon dependency across different temporal intervals and frequencies. Specifically, digitalization intensifies carbon dependency in the short to medium term (below 8 band), while both digitalization and energy transformation significantly reduce carbon dependency in the long term (above 16 band), demonstrating a dynamic correlation among these variables. Furthermore, the results derived from the machine learning tests demonstrate that the influence of digitalization and energy transformation on carbon dependency reveal time-varying effects, digitalization exacerbates carbon dependency within the threshold range of −0.5 to 0.8, whereas energy transformation effectively reduces carbon dependency beyond the threshold of 0.3. This research investigates the complex interrelations among digitalization, energy transformation, and carbon dependency, providing essential experiences and lessons that are applicable to green and sustainable development efforts worldwide. © 2025 Elsevier B.v.

关键词： Carbon dependency Digitalization Energy transformation Machine learning Wavelet analysis

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：