检索结果-内蒙古大学图书馆

International Conference on Innovative Sustainable Computational Technologies (CISCT)

作者： Bhaskaran S Veeramanickam M.R.M S. Hariharan N. Bharathiraja Pradeepa K Raja Marappan School of Computing SASTRA Deemed University Thanjavur India Chitkara University Institute of Engineering and Technology Chitkara University Punjab India Department of Computer Science and Engineering Vardhaman College of Engineering Hyderabad India Department of Computer Science and Engineering St. Michael College of Engineering and Technology India

ISBN: (纸本)9781665474177

Background: Nowadays, pharmaceutical and healthcare industries apply sentiment analysis (SA) to analyze customers' feedback on pharmaceutical products. SA uses natural language processing (NLP) to recognize, assess and analyze the pharmaceutical data to extract valuable insights about the sentiment it conveys. Over the past two decades, in addition to text, emoticons have become a conventional way of expressing one's sentiments. The researchers utilizing textual data for SA have largely ignored emoticons due to their complexity and insufficient resources. Aim of the study: Necessary to include emoticons in SA to capture the emotion expressed by pharmaceutical data. However, this research develops an algorithm that performs SA on text and emoticons. Methods: To highlight the significance of emoticons, the analysis has been conducted on text-only and text-and-emoticon data, using deep learning (DL) and machine learning (ML) methods on multiple datasets. The features – term frequency-inverse document frequency (TF–IDF), emoticon lexicons, and a bag of words (BoW) are taken for simulation. Results: Simulation of text features using BiLSTM provides better accuracy, between 85% and 90%, compared to other methods. The precision lies between 70% and 90%. The recall is lowered to 64%. The sentiment the emoticon conveys outweighs that of the text associated with it. In addition, the DL algorithms outperform the ML methods. The overall results indicate that considering emoticons along with text positively impacts ***, it is noted that the DL model marginally outperforms all of the ML algorithms used. Conclusions: This research analyzed the augmentations to the implementation, including multiple enhancements during the pre-processing stage, the inclusion of emojis, and the usage of a bi-long short-term memory network (BiLSTM) in the DL model. The experimental results conclude that better measures are obtained using the DL method.

关键词： Industries Deep learning Sentiment analysis Analytical models Machine learning algorithms Computational modeling Medical services

来源：评论

学校读者我要写书评

暂无评论

Distilled Transformers with Locally Enhanced Global Representations for Face Forgery Detection

arXiv

引用

arXiv 2024年

作者： Zhang, Yaning Li, Qiufu Yu, Zitong Shen, Linlin Computer Vision Institute College of Computer Science and Software Engineering Shenzhen University Guangdong Shenzhen518060 China National Engineering Laboratory for Big Data System Computing Technology Shenzhen University Guangdong Shenzhen518060 China Shenzhen Institute of Artificial Intelligence and Robotics for Society Guangdong Shenzhen518129 China Guangdong Key Laboratory of Intelligent Information Processing Shenzhen University Guangdong Shenzhen518060 China School of Computing and Information Technology Great Bay University Guangdong Dongguan523000 China

Face forgery detection (FFD) is devoted to detecting the authenticity of face images. Although current CNN-based works achieve outstanding performance in FFD, they are susceptible to capturing local forgery patterns generated by various manipulation methods. Though transformer-based detectors exhibit improvements in modeling global dependencies, they are not good at exploring local forgery artifacts. Hybrid transformer-based networks are designed to capture local and global manipulated traces, but they tend to suffer from the attention collapse issue as the transformer block goes deeper. Besides, soft labels are rarely available. In this paper, we propose a distilled transformer network (DTN) to capture both rich local and global forgery traces and learn general and common representations for different forgery faces. Specifically, we design a mixture of expert (MoE) module to mine various robust forgery embeddings. Moreover, a locally-enhanced vision transformer (LEVT) module is proposed to learn locally-enhanced global representations. We design a lightweight multi-attention scaling (MAS) module to avoid attention collapse, which can be plugged and played in any transformer-based models with only a slight increase in computational costs. In addition, we propose a deepfake self-distillation (DSD) scheme to provide the model with abundant soft label information. Extensive experiments show that the proposed method surpasses the state of the arts on five deepfake datasets. Copyright © 2024, The Authors. All rights reserved.

关键词： Face recognition

来源：评论

学校读者我要写书评

暂无评论

Developing Role Model of PV Powered Battery Swapping Stations for e-scooters in Urban Regions

Developing Role Model of PV Powered Battery Swapping Station...

引用

IEEE International Symposium on Environment Friendly Energies and Applications (EFEA)

作者： Robin Rohde Alexandru Belea Magdalena Seibert Orin Gielen Abdelrahman M. Abdelalim B. Ashok Mohamed Emad Farrag School of Computing Engineering and Built Environment Glasgow Caledonian University Glasgow UK School of Information Technology and Computer Science Nile University Giza Egypt School of Mechanical Engineering Vellore Institute of Technology Vellore Tamil Nadu India

Electric vehicles (EVs) can only provide lower carbon emissions than conventional, internal combustion-powered vehicles if they are charged using green energy. They also have the drawback of long charging times to “refuel” them. To combat these two problems, a solar-powered battery charging and swapping station was developed using centered-human design and systems engineering. The design focuses on electric two-wheeled vehicles (e-scooters) due to the easier handling of their light and low-capacity batteries compared to electric cars. These stations are easy to maintain and manage and can be run by low-income individuals to provide them with a business opportunity. The stations are designed to run independently from the electrical grid and thus be deployed in areas with less sophisticated infrastructure to achieve lower carbon emissions. A based-Mobile application was also designed to identify the next available station with the same battery rating. The project outcome is found reliable in developing economically, ecologically, and socially sustainable services for most needing communities.

关键词： Franchising Renewable energy sources Costs Urban areas Focusing Carbon dioxide Batteries

来源：评论

学校读者我要写书评

暂无评论

Heterogeneous Face Recognition Algorithm: Convolution Neural Network Approach

Heterogeneous Face Recognition Algorithm: Convolution Neural...

引用

Intelligent Communication Technologies and Virtual Mobile Networks (ICICV), International Conference on

作者： E. Sujatha M. Manickam. R.Ani Minisha K.S. Rekha T. Indumathy Department of Computer Science and Engineering Saveetha Engineering College (Autonomous) Chennai Tamil Nadu Department of Networking and Communications School of Computing SRM Institute of Science and Technology Chennai Tamil Nadu Department of Artificial Intelligence and Data Science Vel Tech Multi Tech Dr. RR and Dr SR Engineering College Tamil Nadu Department of Medical Electronics Saveetha Engineering College (Autonomous) Chennai Tamil Nadu

Heterogeneous face recognition, also known as HFR, is an approach to facial recognition that attempts to match facial pictures that have been gathered through a variety of sensory modalities. HFR is primarily used for forensics, network security, and all commercial applications. The heterogeneous face recognition (HFR) algorithm is robust in handling challenges that exist in conventional face recognition. This is since heterogeneous face images display an enormous amount of intra- class variation, and it does not support hybrid modality of face images. This study provides a unique technique to transfer learning as a means of mastering invariant facial feature characteristics in visual (VIS) face pictures. A wrapping procedure combines end to end connections of two components for deep network. A discriminator has the details of fine-grained data for improving the quality of image which depends on wavelets. The system is made up of two different modules: one deals with face recognition, while the other determines a person's gender. It analysis about the characteristics of gender from the image using Convolutional Neural Networks that have been pre-trained. To train CNN in the face recognition module, in particular, the use of the open-source datasets Labelled Faces within the Wild, YouTube Face, and VGGFace2 results in an improvement in accuracy. Deep learning techniques and a substantial number of photographic examples of people's faces are what make it feasible for the proposed recognition to function so admirably. The problem of studying recognition remains challenging due to the limited size of the sample and the information gap between the senses.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A Multilevel Approach for Solving Large-Scale QUBO Problems with Noisy Hybrid Quantum Approximate Optimization

A Multilevel Approach for Solving Large-Scale QUBO Problems ...

引用

IEEE Conference on High Performance Extreme computing (HPEC)

作者： Filip B. Maciejewski Bao G. Bach Maxime Dupont P. Aaron Lott Bhuvanesh Sundar David E. Bernal Neira Ilya Safro Davide Venturelli Quantum Artificial Intelligence Laboratory (QuAIL) NASA Ames Research Center CA USA USRA Research Institute for Advanced Computer Science (RIACS) CA USA Department of Computer and Information Sciences University of Delaware Newark DE USA Rigetti Computing Berkeley CA Davidson School of Chemical Engineering Purdue University West Lafayette IN USA

ISBN: (数字)9798350387131

ISBN: (纸本)9798350387148

Quantum approximate optimization is one of the promising candidates for useful quantum computation, particularly in the context of finding approximate solutions to Quadratic Unconstrained Binary Optimization (QUBO) problems. However, the existing quantum processing units (QPUs) are of relatively small size, and canonical mappings of QUBO via the Ising model require one qubit per variable, rendering direct large-scale optimization infeasible. In classical optimization, a general strategy for addressing many large-scale problems is via multilevel/multigrid methods, where the large target problem is iteratively coarsened and the global solution is constructed from multiple small-scale optimization runs. In this work, we experimentally test how existing QPUs perform when used as a sub-solver within such a multilevel strategy. To this aim, we combine and extend (via additional classical processing steps) the recently proposed Noise-Directed Adaptive Remapping (NDAR) and Quantum Relax & Round (QRR) algorithms. We first demonstrate the effectiveness of our heuristic extensions on Rigetti's superconducting transmon device Ankaa-2. We find approximate solutions to 10 instances of fully connected 82-qubit Sherrington-Kirkpatrick graphs with random integer-valued coefficients obtaining normalized approximation ratios (ARs) in the range $\sim 0.98-1.0$ , and the same class with real-valued coefficients (ARs $\sim 0.94-1.0)$ . Then, we implement the extended NDAR and QRR algorithms as subsolvers in the multilevel algorithm for 6 large-scale graphs with at most $\sim 27,000$ variables. In practice, the QPU (with classical post-processing steps) is used to find approximate solutions to dozens of at most 82-qubit problems, which are iteratively used to construct the global solution. We observe that quantum optimization results are competitive in terms of the quality of solutions when compared to classical heuristics used as subsolvers within the multilevel approach. Reproducibility

关键词： Superconducting devices Performance evaluation Quantum advantage Source coding Qubit Pipelines Approximation algorithms Rendering (computer graphics) Noise measurement Optimization

来源：评论

学校读者我要写书评

暂无评论

Recommendation System using Inference-based Graph Learning – Modeling and Analysis

Recommendation System using Inference-based Graph Learning ...

引用

International Conference on Innovative Sustainable Computational Technologies (CISCT)

作者： Bhaskaran S. S. Hariharan Veeramanickam M.R. M N. Bharathiraja Pradeepa K Raja Marappan School of Computing SASTRA Deemed University Thanjavur India Department of Computer Science and Engineering Vardhaman College of Engineering Hyderabad India Chitkara University Institute of Engineering and Technology Chitkara University Punjab India Department of Computer Science and Engineering St. Michael College of Engineering and Technology India

ISBN: (纸本)9781665474177

Recently, the development of recommendation systems has been an emerging area where the target is the learner's preferences, and styles should be learned from past interactions to predict the ratings and recommendations of the new items. The significant critiques and challenges are some of the datasets are sparse and the bias in the ratings and popularity of the items that conventional recommenders can't efficiently process. This research constructs the graph model to merge the items and user interaction to enhance user representations. The new inference operator is designed to remove the bias factors and to draw inferences. The users and items are represented with high quality by constructing the semantic space that connects the users and the item's characteristics. The proposed model is simulated on real-world datasets, proving that the model produces significant outcomes better than existing methods.

关键词： Analytical models Computational modeling Semantics Predictive models Recommender systems

来源：评论

学校读者我要写书评

暂无评论

Application of Deep Convolutional Neural Networks (DCNNs) for Automated Classification of E-Waste Components: Focusing on Circuit Boards, Batteries, and Cables

Application of Deep Convolutional Neural Networks (DCNNs) fo...

引用

Innovations in High Speed Communication and Signal Processing (IHCSP), International Conference on

作者： Frederick Ruby Helen Sivaram Rajeyyagari Murugananth Gopal Raj Shamim Ahmad khan Varunadevi S I Infant raj Department of Humanities and Social Sciences Saveetha School of Law Saveetha Institute of Medical and Technical Sciences Chennai India Department of Computer Science College of Computing and Information Technology Shaqra University Shaqra Saudi Arabia Department of EEE Ahalia School of Engineering and Technology Palakkad Kerala India Glocal School of Science & Technology Glocal University U.P India Department of CSE CMS College of Engineering and Technology Coimbatore Tamilnadu Department of CSE K.Ramakrishnan College of Engineering Trichy Tamil Nadu India

ISBN: (数字)9798350368949

ISBN: (纸本)9798350368956

The proliferation of electronic waste (e-waste) has become a pressing environmental concern, necessitating efficient methods for its classification and management. Though there are traditional methods which are used when sorting and recyling e-waste, they are time-consuming, costly and most of the time, are not accurate and hence, a lot of resources are wasted in the process. This research aims at extending the use of Deep Convolutional Neural Networks (DCNNs) in the classification of critical e-waste components namely circuit boards, batteries, and cables. The proposed method fully utilizes the capacity of extracting features of DCNNs in order to accurately detect and sort these components by means of the receiving visual information. For training and testing the DCNN model a dataset was created and preprocessed and the target images included circuit boards of different orientations and sizes, batteries and cables. The architecture of a network was created considering aspect of depth and computational complexity guarantying high accuracy and feasible time for great data set processing. In the evaluation process of the model, a comparison was made with various standard performance measures, with checklists such as accuracy, precision and recall and F1-score to rate the prowess of the proposed model in taking part in the classifying process of e-waste components. From the analysis, the proposed DCNN model reached classification accuracy over 95.9% on all the categories suggesting the model can be used as a tool to sort e-waste automatically. The paper also considers the issues of implementing such technology in the current recycling systems and the overall advantages, which will be the increase of the sorting efficiency, decreased labor expenses, as well as elevated recovery rates of materials demanded in the market.

关键词： Accuracy Computational modeling Printed circuits Electronic waste Feature extraction Batteries Convolutional neural networks Integrated circuit modeling Cables Sorting

来源：评论

学校读者我要写书评

暂无评论

A Virtual Reality System for Predictive Display Functionality in a Telexistence-Controlled SEED-Noid Humanoid Robot with Evaluation of VR Sickness

A Virtual Reality System for Predictive Display Functionalit...

引用

IEEE International Workshop on Robot and Human Communication (ROMAN)

作者： Taiga Suda Motoki Yodowatari Sosuke Kosaki Koki Yokoyama Eito Yanagisawa Eimei Oyama Kohei Tokoi Hiroyuki Okada Arvin Agah Sousuke Nakamura Graduate School of Science and Engineering Hosei University Japan National Institute of Advanced Industrial Science and Technology Japan College of Engineering Tamagawa University Machida City Tokyo JAPAN Wakayama University Wakayama City Wakayama JAPAN Electrical Engineering and Computer Science The University of Kansas Lawrence KS USA Faculty of Science and Engineering Hosei University Japan

Telexistence technology enables the manipulation of a remote robot using sensory feedback and leader-follower control, giving the operator the impression of controlling their own body. However, the robot’s operability declines significantly due to increasing communication time delays. In order to address this issue, the "Predictive Display" technique has been proposed and developed. This technique displays both the real robot’s movements with time delay and the computer-generated (CG) robot’s movements without time delay. In the Telexistence robot operation system, a predictive display system can be constructed using a virtual reality (VR) system. In this study, we have developed a VR system for Telexistence operation with THK’s SEED-Noid humanoid robot. Unity is utilized, which enables the predictive display function for the stereo camera mounted on the SEED-Noid. We developed and evaluated a display mode without neck joint angle restrictions to reduce VR sickness and workload using the developed VR system. Additionally, we developed a display mode with a limited neck joint angle for comparison purposes.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Application of Deep Convolutional Neural Networks (DCNNs) for Automated Classification of E-Waste Components: Focusing on Circuit Boards, Batteries, and Cables 2

Application of Deep Convolutional Neural Networks (DCNNs) fo...

引用

2nd IEEE International Conference on Innovations in High-Speed Communication and Signal Processing, IHCSP 2024

作者： Helen, Frederick Ruby Rajeyyagari, Sivaram Raj, Murugananth Gopal Khan, Shamim Ahmad Varunadevi, S. Raj, I. Infant Saveetha Institute of Medical and Technical Sciences Saveetha School of Law Department of Humanities and Social Sciences Chennai India Shaqra University College of Computing and Information Technology Department of Computer Science Shaqra Saudi Arabia Ahalia School of Engineering and Technology Department of Eee Kerala Palakkad India Glocal University Glocal School of Science & Technology U.P India Cms College of Engineering and Technology Department of Cse Tamilnadu Coimbatore India K.Ramakrishnan College of Engineering Department of Cse Tamil Nadu Trichy India

ISBN: (纸本)9798350368949

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Sports-QA: A Large-Scale Video Question Answering Benchmark for Complex and Professional Sports

arXiv

引用

arXiv 2024年

作者： Li, Haopeng Deng, Andong Liu, Jun Rahmani, Hossein Guo, Yulan Schiele, Bernt Bennamoun, Mohammed Ke, Qiuhong School of Computing and Information Systems University of Melbourne Australia Center for Research in Computer Vision University of Central Florida United States Pillar Singapore University of Technology and Design Singapore School of Computing and Communications Lancaster University United Kingdom School of Electronics and Communication Engineering Sun Yat-sen University China Department of Computer Vision and Machine Learning Max Planck Institute for Informatics Saarland Informatics Campus Germany School of Physics Maths and Computing University of Western Australia Australia Department of Data Science & AI Monash University Australia

Reasoning over sports videos for question answering is an important task with numerous applications, such as player training and information retrieval. However, this task has not been explored due to the lack of relevant datasets and the challenging nature it presents. Most datasets for video question answering (VideoQA) focus mainly on general and coarse-grained understanding of daily-life videos, which is not applicable to sports scenarios requiring professional action understanding and fine-grained motion analysis. In this paper, we introduce the first dataset, named Sports-QA, specifically designed for the sports VideoQA task. The Sports-QA dataset includes various types of questions, such as descriptions, chronologies, causalities, and counterfactual conditions, covering multiple sports. Furthermore, to address the characteristics of the sports VideoQA task, we propose a new Auto-Focus Transformer (AFT) capable of automatically focusing on particular scales of temporal information for question answering. We conduct extensive experiments on Sports-QA, including baseline studies and the evaluation of different methods. The results demonstrate that our AFT achieves state-of-the-art performance. Copyright © 2024, The Authors. All rights reserved.

关键词： Question answering

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：