检索结果-内蒙古大学图书馆

International Conference on Advanced computing and Communication Systems (ICACCS)

作者： Bandi Rupendra Reddy Daka Chandra Rup Mathi Rohith Meena Belwal Department of Computer Science and Engineering Amrita School of Computing Bengaluru India

Communication is the key behind the whole human evolution, and there would be nothing possible if the stream for communicating our thoughts to one another were cut off. Language is a significant factor when it comes to conveying one's ideas, thoughts, and information to other people. Sign language is used to communicate hard-of-hearing people's opinions. Every country has different sign languages. In this paper, we have used ISL (Indian Sign Language). To ease communication between ordinary people and deaf/dumb people, we propose the design and implementation of a model that translates a live voice, audio recordings, or text of a native Indian regional language (Tamil) to text and then further matches it to sign language animations from the video animation dataset. The speech is converted to text using two deep learning models LSTM(Long Short Term Memory), Bi-LSTM, and Google API. Then the text is transformed into a sign using ISL (Indian Sign Language) dataset. The proposed models achieved 45%, 65%, and 95% accuracy for LSTM, Bi-LSTM, and Google API, respectively.

关键词： Deep learning Computational modeling Gesture recognition Assistive technologies Animation Audio recording Internet

来源：评论

学校读者我要写书评

暂无评论

Astronomical Image Clustering Using Spectral, K-Means and Mean Shift

Astronomical Image Clustering Using Spectral, K-Means and Me...

引用

2025 International Conference on computing for Sustainability and Intelligent Future, COMP-SIF 2025

作者： Mohite, Vaishnavi Shivaji Patra, Krishna Gayatri Sankar, C.H. Hari Jayan, Sarada Amrita School of Computing Department of Computer Science Bengaluru India Amrita School of Engineering Department of Mathematics Bengaluru India

ISBN: (纸本)9798331538538

By analyzing vast amounts of observational data through astronomical image analysis, you can gain valuable information about the structure of the universe. This paper presents a new method for clustering astronomical images that uses both spectral and K-means methods and also incorporates linear algebra and optimization principles. We propose a method that combines feature extraction techniques such as the Fourier transform with clustering algorithms to classify celestial objects based on their characteristics. The research examines the effectiveness of spectral clustering and K-means clustering in organizing astronomical images into meaningful groups, taking into account factors such as computer efficiency, clustering accuracy and interpretability. Experimenting with different datasets, including telescope images, we demonstrate the effectiveness of our approach in automating the classification of astronomical objects. Our results reveal insights into the underlying structure of astronomical knowledge and offer promising avenues for future research in computational astronomy. In general, this research advances both the field of astronomy and the application of mathematical methods to image analysis and clustering. © 2025 IEEE.

关键词： K-means clustering

来源：评论

学校读者我要写书评

暂无评论

Graph-Based Generation and Validation of Use Case Diagrams

引用

Procedia computer science 2025年 259卷 1356-1365页

作者： Reshma P Nair M G Thushara K Somasundaram Department of Computer Science and Engineering Amrita Vishwa Vidyapeetham Amrita School of Computing Amritapuri Kerala India Department of Computer Science and Applications Amrita Vishwa Vidyapeetham Amrita School of Computing Amritapuri Kerala India Department of Mathematics Amrita School of Physical Sciences Amrita Vishwa Vidyapeetham Coimbatore India

As we know in today’s modern software engineering, use case diagrams for representing the functionalities of the system act significantly in mediating the gap between the stakeholder and the developer. Some of the earlier techniques used in constructing such diagrams can be problematic when these diagrams are to be employed in large systems where there are many actors and activities. That is why every limitation outlined above might be resolved further with the help of NLP and developments in graphs theory. The proposed method makes improvements by reducing the human interference; it can handle increase in system size and the number of actors and their interactions; it also provides highly accurate shortest-path identification through use of validation methods. In addition, shortest-path algorithms are incorporated to improve the validation exercise, thereby minimizing errors common in presenting the system functionalities. Hence, the results indicate that the proposed method is more effective not only for generating use case diagrams with increased precision and scalability but also for constructing the necessary framework for checking system interactions’ correctness. This work provides a foundation for future studies aimed at enhancing the automation of the use case diagram especially for complex systems.

关键词： NLP Graph Theory Use case Diagrams Automatic Generation Shortest-Path Algorithms

来源：评论

学校读者我要写书评

暂无评论

Transformer-Enhanced U-Net with Transfer Learning for Accurate Detection of Eye Coloboma

Transformer-Enhanced U-Net with Transfer Learning for Accura...

引用

International Conference on Power, Energy, Control and Transmission Systems (ICPECTS)

作者： A Dhayanithi K. Bala Department of Computer Science and Engineering School of Computing Bharath Institute of Higher Education and Research Chennai Tamil Nadu India Department of Computer Science and Engineering School of Computing Bharath Institute of Higher Education and Research Chennai Tamil Nadu

ISBN: (数字)9798331508845

ISBN: (纸本)9798331508852

Eye coloboma is a type of congenital eye abnormality that causes developmental abnormalities of the eye which may culminate in eye disorders that fundamentally derange eyesight. Fractional and timely diagnosis remains highly important to its proper managing and treating. This work proposes a model with a Transformer-Enhanced U-Net design implemented with Transfer Learning to enhance the ability of the model to detect and segment eye coloboma from medical imaging data. This work builds upon the benefits of U-Net which has shown promising results especially in the biomedical image segmentation problem through addition of transformer modules that introduce long range dependencies and contextual information. This enhancement makes helps the model to pay more attention to such minor aspects as coloboma and general diagnostic accuracy consequently. Thanks to Transfer Learning with a pretrained backbone, for example Efficient Net the system is provided with Feature Representations recovered from large datasets that are already available and training time is drastically cut short while the performance of the system is not compromised at *** approach includes fine-tuning of Transformer-Enhanced U-Net model using a selected set of eye images, labelled for coloboma existence. By including the self-attention mechanism, the model is capable of focusing on the critical regions in the image, increasing its responsiveness to any form of coloboma. Preliminary tests show that this scheme performs better than the standard convolutional networks for segmenting complex regions by minimizing false negatives. Further, the model’s flexibility to different imaging conditions demonstrates its suitability in realistic clinical applications. The objective of this research is to come up with a dependable, Auto-Generated diagnostic tool that will help the ophthalmologists in the early diagnosis of the eye related illnesses hence making increased positive outcomes for eye patients. The Transf

关键词： Training Image segmentation Accuracy Sensitivity Biological system modeling Transfer learning Transformers Data models Standards Biomedical imaging

来源：评论

学校读者我要写书评

暂无评论

AST-Based and Token-Based Neural Networks for Source Code Classification: A Comparative Performance Analysis

AST-Based and Token-Based Neural Networks for Source Code Cl...

引用

International Conference on computing and Networking Technology (ICCNT)

作者： R Parvathy Mg Thushara Department of Computer Science & Engineering Amrita School of Computing Amrita vishwa Vidyapeetham Amritapuri India Department of Computer Science & Applications Amrita School of Computing Amrita vishwa Vidyapeetham Amritapuri India

ISBN: (数字)9798350370249

ISBN: (纸本)9798350370270

Programming languages function as a medium for expressing instructions in the context of computer program creation. A prevalent convention in programming is categorizing source code into fragments, a strategy that enhances understanding and simplifies the upkeep process. This research study presents a comparative examination of two distinct models utilized to classify source code. The first model employs a neural network approach that heavily relies on Syntax Trees (ASTs). In contrast, the second model is built on a neural network framework that utilizes the tokenization process. The AST-based model employs a methodology that entails decomposing extensive ASTs into smaller clusters of statement trees. These clusters are subsequently transformed into vectors by capturing the lexical and syntactical characteristics of the statements. The naturalness of certain statements is captured by employing a bidirectional Recurrent Neural Network (RNN) model, which utilizes a collection of statement vectors. In contrast, the token-based paradigm diverges from ASTs and emphasizes tokens obtained through tokenization. The tokens undergo preprocessing and vectorization before inputting into a bidirectional RNN model to evaluate the naturalness of the statements. To choose the optimal model, the accuracy of the two models are compared.

关键词： computer languages Recurrent neural networks Codes Computational modeling Source coding Data preprocessing Syntactics Vectors Tokenization Performance analysis

来源：评论

学校读者我要写书评

暂无评论

A Study of Search User Interface Design based on Hofstede's Six Cultural Dimensions 6

A Study of Search User Interface Design based on Hofstede's ...

引用

6th International Conference on computer-Human Interaction Research and Applications, CHIRA 2022

作者： Chessum, Karen Liu, Haiming Frommholz, Ingo School of Computer Science and Technology University of Bedfordshire Luton United Kingdom School of Electronics and Computer Science University of Southampton United Kingdom School of Engineering Computing and Mathematical Sciences University of Wolverhampton Wolverhampton United Kingdom

ISBN: (纸本)9789897586095

An information seeker's cultural background could influence their preference for search user interface (UI) design. To study cultural influences Geert Hofstede's cultural dimensions have been applied to website design for a number of years. In this paper, we examine if Hofstede's six cultural dimension can be applied to inform the design of search engine user interfaces. The culturally designed search user interfaces have been evaluated in a study with 148 participants of different cultural backgrounds. The results have been analysed to determine if Hofstede's cultural dimensions are appropriate for understanding users' preferences on search user interface design. Whilst the key findings from the study suggest Hofstede cross-cultural dimensions can be used to model users' preferences on search interface design, further work is still needed for particular cultural dimensions to reinforce the conclusions. Copyright © 2022 by SCITEPRESS - science and Technology Publications, Lda. All rights reserved.

关键词： Web Design

来源：评论

学校读者我要写书评

暂无评论

Safety and Completeness in Flow Decompositions for RNA Assembly 26th

Safety and Completeness in Flow Decompositions for RNA As...

引用

26th International Conference on Research in Computational Molecular Biology, RECOMB 2022

作者： Khan, Shahbaz Kortelainen, Milla Cáceres, Manuel Williams, Lucia Tomescu, Alexandru I. Department of Computer Science and Engineering IIT Roorkee Roorkee India Department of Computer Science University of Helsinki Helsinki Finland School of Computing Montana State University Bozeman United States

ISBN: (纸本)9783031047480

Flow decomposition has numerous applications, ranging from networking to bioinformatics. Some applications require any valid decomposition that optimizes some property as number of paths, robustness, or path lengths. Many bioinformatic applications require the specific decomposition which relates to the underlying data that generated the flow. Thus, no optimization criteria guarantees to identify the correct decomposition for real inputs. We propose to instead report the safe paths, which are subpaths of at least one path in every flow decomposition. Ma et al. [WABI 2020] addressed the existence of multiple optimal solutions in a probabilistic framework, which is referred to as non-identifiability. Later, they gave a quadratic-time algorithm [RECOMB 2021] based on a global criterion for solving a problem called AND-Quant, which generalizes the problem of reporting whether a given path is safe. We present the first local characterization of safe paths for flow decompositions in directed acyclic graphs, giving a practical algorithm for finding the complete set of safe paths. We also evaluated our algorithm against the trivial safe algorithms (unitigs, extended unitigs) and a popular heuristic (greedy-width) for flow decomposition on RNA transcripts datasets. Despite maintaining perfect precision our algorithm reports ≈ 50% higher coverage over trivial safe algorithms. Though greedy-width reports better coverage, it has significantly lower precision on complex graphs. On a unified metric (F-Score) of coverage and precision, our algorithm outperforms greedy-width by ≈ 20%, when the evaluated dataset has significant number of complex graphs. Also, it has superior time (3–5 × ) and space efficiency (1.2–2.2 × ), resulting in a better and more practical approach for bioinformatics applications of flow decomposition. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： RNA

来源：评论

学校读者我要写书评

暂无评论

Deep Reinforcement Learning for natural language understanding and Dialogue Systems

Deep Reinforcement Learning for natural language understandi...

引用

International Conference on Recent Trends in Advance computing (ICRTAC)

作者： Prasanna Kumar R Venkatraman K Siva Jyothi Natha Reddy B V V V Bhagya Sree Vemireddy Anvitha Manojna K Sudarshan R Computer Science and Engineering Amrita Vishwa Vidyapeetham School of Computing Chennai

Innovations in neural dialogue models promise to improve conversational AI. However, these models frequently lack foresight, handling answers one by one without considering long-term consequences. Traditional NLP techniques included reinforcement learning to address this. Deep reinforcement learning predicts future rewards in chatbot talks, according to this study. The model simulates agent interactions using policy gradient approaches, prioritizing sequences with essential conversational characteristics such as informativeness and coherence. Variety, answer length, and human assessment are emphasized in the evaluation of dialogue simulations, demonstrating the potential for interesting, protracted talks. This study is the first step toward developing a neural conversational model that prioritizes long-term discourse success.

关键词：

来源：评论

学校读者我要写书评

暂无评论

BrainCrossFed CNN Model for Alzheimer Classification using MRI data and Comparison and Benchmarking proposed model with DINOv2 and ExplainableAI using GradCAM

BrainCrossFed CNN Model for Alzheimer Classification using M...

引用

2023 International Conference on the Confluence of Advancements in Robotics, Vision and Interdisciplinary Technology Management, IC-RVITM 2023

作者： Rashmi, Uppin Beena, B.M. Ambesange, Sateesh Amrita Vishwa Vidyapeetham Amrita School of Computing Cse Dept Bangalore India Amrita Vishwa Vidyapeetham Amrita School of Computing Department of Computer Science & Engineering Bangalore India Bangalore Pragyan SmartAI Technology Llp Pragyan AI Pragyan Ai School India

ISBN: (纸本)9798350306088

Alzheimer is one of the key mental illnesses suffered by 55 million people worldwide and 10 million new patients are added every year. If Alzheimer's is detected in early stage, then the impact of disease will be reduced. Artificial intelligence (AI) models are helping in early diagnosis using MRI images. For better AI model development, the model needs to be trained over a huge dataset. But Health data privacy is another key concern of data collection, and hence federated learning framework helps models get trained over data without data transferring to cloud platform. Other Key concerns of AI model development is the availability of labeled data and benchmarking the model performance developed on dataset, which is trimmed or is a modified version. As availability of research work on specific data is rare. Currently, researchers have to develop other research models for specific data and such approaches take more time. Also, doctors expect explainability of the model, instead of black box model. Proposed BrainCrossFed model addresses data privacy concerns by avoiding the need for data transfer to cloud and leverage available labeled data at different federated nodes along with a dataset for Alzheimer to compute cross fed average of model weights. Due to the proposed Cross Fed Avg algorithm performance of the proposed deep learning model accuracy is increased from 99.11% to 99.77% close to 100%. Also such cross fed average enables the model to find global minima. Final performance achieved by the proposed model is 99.77%, with 100%. The performance of the model is benchmarked with current well known classification models, State-of-the-Art (SOTA), self supervised models like DINOv2 [1], for the same datasets. DINOv2 model has achieved 98%. Other generic well known models like ResNet101, DenseNet121, and Visual Geometry Group16 (VGG16) and customer models [2] have achieved max performance 97.60%. The GradCAM developed to explain the classification result. © 2023 IEEE.

关键词： Magnetic resonance imaging

来源：评论

学校读者我要写书评

暂无评论

3D Image Reconstruction and Processing for Augmented and Virtual Reality Applications: A computer Generated Environment 2

3D Image Reconstruction and Processing for Augmented and Vir...

引用

2nd IEEE International Conference on Communication, Security and Artificial Intelligence, ICCSAI 2023

作者： Baskar, Radhika Krishnammal, P. Muthu Aeron, Anurag Ali, Syed Sumera Leonid, T. Thomas Arun, M.R. Saveetha School of Engineering Saveetha Institute of Medical and Technical Sciences Chennai India School of Electronics Engineering VIT-AP University Andhra Pradesh Amaravati India School of Computing Dit University Computer Science and Engineering Dehradun India Csmss Chh Shahu College of Engineering Electronics and Computer Engineering Department Maharashtra Aurangabad India Ece Dept Kcg College of Technology Chennai India Vel Tech Rangarajan Dr. Sagunthala R&D Science and Technology Department of Ece Chennai India

ISBN: (纸本)9798350369960

New technologies like AR and VR are changing our online and offline experiences. Augmented and virtual reality applications need 3D picture reconstruction and processing. In augmented and virtual reality, this abstract explains 3D image reconstruction and processing concepts, techniques, and effects. Creating engaging AR and VR experiences starts with 3D data. Depth sensors, stereoscopic images, and photogrammetry are essential for accurate 3D reconstructions of real-world objects and settings. These data streams support spatial cognition and interaction in AR/VR. Reconstruction technologies like point cloud and mesh reconstruction are essential for turning 3D data into models. Virtual features that complement or replace the actual environment in AR experiences or build fully fabricated VR settings are based on these models. By applying 2D textures to 3D geometry, 'texture mapping' enhances scene realism. Realtime augmented and virtual reality processing requires effective rendering, tracking, and physics simulation algorithms. Rendering engines provide beautiful, responsive experiences, while SLAM lets users explore and interact with virtual information in real life. Occlusion culling and level-of-detail rendering are two optimisation methods needed for smooth performance on high-end VR systems and mobile AR devices. Virtual reality is defined by 3D object interaction and control. Natural and engaging interfaces are created using gesture detection and real-time physics simulations. AR and VR software developers encounter new challenges with each iteration. AR requires accurate spatial mapping, dependable tracking, and simple user interfaces to blend the virtual and real worlds. In contrast, virtual reality (VR) immerses users in realistic computer settings. Portability across operating systems is crucial for greater adoption. Developers must optimise programmes for several hardware platforms to reach the most people. © 2023 IEEE.

关键词： 3D image Augmented reality Reconstruction Virtual Reality

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：