检索结果-内蒙古大学图书馆

arXiv 2022年

作者： Hao, Yanbin Wang, Shuo Cao, Pei Gao, Xinjian Xu, Tong Wu, Jinmeng He, Xiangnan The CCCD Key Lab of Ministry of Culture and Tourism School of Data Science School of Information Science and Technology University of Science and Technology of China Anhui 230026 China The Wuhan Research Institute of Posts and Telecommunications Hubei Wuhan430205 China The School of Computer Science and Information Engineering School of Artificial Intelligence Hefei University of Technology Anhui 230009 China The School of Data Science School of Computer Science and Technology University of Science and Technology of China Anhui 230026 China The Hubei Key Laboratory of Optical Information and Pattern Recognition Wuhan Institute of Technology Hubei Wuhan430070 China

Attention mechanisms have significantly boosted the performance of video classification neural networks thanks to the utilization of perspective contexts. However, the current research on video attention generally focuses on adopting a specific aspect of contexts (e.g., channel, spatial/temporal, or global context) to refine the features and neglects their underlying correlation when computing attentions. This leads to incomplete context utilization and hence bears the weakness of limited performance improvement. To tackle the problem, this paper proposes an efficient attention-in-attention (AIA) method for element-wise feature refinement, which investigates the feasibility of inserting the channel context into the spatio-temporal attention learning module, referred to as CinST, and also its reverse variant, referred to as STinC. Specifically, we instantiate the video feature contexts as dynamics aggregated along a specific axis with global average and max pooling operations. The workflow of an AIA module is that the first attention block uses one kind of context information to guide the gating weights calculation of the second attention that targets at the other context. Moreover, all the computational operations in attention units act on the pooled dimension, which results in quite few computational cost increase ( Copyright © 2022, The Authors. All rights reserved.

关键词： Machine learning

来源：评论

学校读者我要写书评

暂无评论

Leveraging LSTM-based Auto-Encoders for Covid-19 Severity Prediction and Classification

Leveraging LSTM-based Auto-Encoders for Covid-19 Severity Pr...

引用

Recent Advances in science and engineering technology (ICRASET), International Conference on

作者： Supriya Rai Komal Madhukar Birare K.S. Bhuvaneshwari Tarang Bhatnagar S. Jayasree Ashu Katyal School of Mangement - UG Management JAIN (Deemed to be University) Bangalore Karnataka India IT Vishwakarma Institute of Technology Pune India Department of Artificial Intelligence Data science Karpagam College of Engineering Coimbatore India Centre of Research Impact and Outcome Chitkara University Rajpura Punjab India Department of S&H Prince Shri Venkateshwara Padmavathy Engineering College Chennai India Chitkara Centre for Research and Development Chitkara University Himachal Pradesh India

ISBN: (数字)9798350388602

ISBN: (纸本)9798350388619

The primary purpose of this study activity is to give its share to the battle against the growing threats posed by COVID-19 all over the world, which has over one million cases. According to the World Health Organisation (WHO), more than seven million individuals have been reported to have passed away as a result of; the work is focused on building an optimization-based deep learning model that will assist in identifying COVID-19 patients based on early clinical information. The suggested model is pretty accurate for the prediction of COVID-19 and improves the strictness of the forecasts of the ailment. This is accomplished by combining the medical and personal data from a variety of populations. In the preliminary stage, the only things that can be determined are the pre-processing phase and the characteristics that are pertinent for classification using ESDO. The efficacy is characterised by the use of a classifier that combines LSTM and AE. This allows for the determination of the degree of severity, which may range from mild to moderate to severe, with a precision of $\mathbf{9 9 \%} \mathbf{. 0. 0 2}$, and a level of accuracy of $\mathbf{9 8}$. Therefore, future study will concentrate on applying the offered approach for the affected regions in India exclusively, and efforts will also be made to compare the suggested methodology to larger databases, if any such databases exist.

关键词： COVID-19 Training Deep learning Humanities Accuracy databases Predictive models Feature extraction Long short term memory Optimization

来源：评论

学校读者我要写书评

暂无评论

Methodology for Real Estate Recommendation Based on Customer Behavior Knowledge as Context-Approach 15th

Methodology for Real Estate Recommendation Based on Customer...

引用

15th International Joint Conference on Knowledge Discovery, Knowledge engineering and Knowledge Management, IC3K 2023

作者： Vu, Sang Nguyen, Tinh Nguyen, Truc Pham, Vuong T. Huynh, Tai Nguyen, Hien D. University of Information Technology Ho Chi Minh City Viet Nam Faculty of Mathematics and Computer Science University of Science Ho Chi Minh City Viet Nam Vietnam National University Ho Chi Minh City Viet Nam Institute of Data Science and Artificial Intelligence Sai Gon University Ho Chi Minh City Viet Nam Kyanon Digital Ho Chi Minh City Viet Nam Computer Science Department New Mexico State University Las Cruces United States

ISBN: (纸本)9783031875687

Real estate recommendation systems present a significant challenge due to the multifaceted nature of buyer decision-making. This growing field necessitates the development of a holistic recommendation approach. This study proposes the Est-Model, which captures customer behavior related to real estate purchase decisions. The Est-Model leverages a relational knowledge model to organize property features. This model is then integrated with a database containing user profiles and real estate information. To address the cold-start problem for both items and users, the study introduces a novel real estate recommendation methodology. Utilizing a session-based recommendation framework, the system effectively manages sequential and contextual data to predict user preferences for future interactions. Empirical evaluations demonstrate the superiority of this approach over existing methodologies within the top-n recommendation paradigm. These findings are validated through experiments conducted on a real estate dataset. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Knowledge acquisition

来源：评论

学校读者我要写书评

暂无评论

Navigating Grid Dynamics with V2G technology and Naive Bayes Optimization in IoT Systems

Navigating Grid Dynamics with V2G Technology and Naive Bayes...

引用

Innovation in technology (INOCON), IEEE International Conference for

作者： M. Arulprakash Ramakrishnan Raman Atul A Gokhale K. Saravanan Ishwarya M. V S. Sujatha Department of Computing Technologies SRM Institute of Science and Technology Kattankulathur Chennai Tamil Nadu India Symbiosis Institute of Business Management Pune & Symbiosis International (Deemed University) Pune Maharashtra India Symbiosis School of Culinary Arts Symbiosis International (Deemed University) Pune Maharashtra India Department of Information Technology R.M.D. Engineering College Chennai Tamil Nadu India Artificial Intelligence and Data Science Department Agni College of Technology Chennai Tamil Nadu India Department of Biomedical Engineering Saveetha School of Engineering Saveetha Institute of Medical and Technical Sciences Saveetha University Chennai Tamil Nadu India

Vehicle-to-Grid (V2G) technology and Naive Bayes optimization, which might help IoT devices, navigate grid dynamics. Electric vehicles (EVs) with bidirectional charging capability may provide grid assistance using their energy storage capacity. These cars can take and return energy from the grid using V2G technology, stabilizing the system. This paper proposes a unique technique to intelligently control energy flow between the grid and EVs using V2G technology and Naive Bayes optimization algorithms. Using previous data on energy consumption, grid circumstances, and vehicle charging behavior, the Naive Bayes algorithm makes real-time predictions and decisions. In IoT-enabled smart grids, this connection improves energy management efficiency and reliability. The implementation methods and experimental findings of the suggested strategy are discussed in the study. The system grid stability, energy usage, and grid dynamics management are also assessed. The results show that V2G technology and Naive Bayes optimization in IoT systems can intelligently navigate grid dynamics, creating a more sustainable and resilient energy infrastructure. To smart grid expertise and shows how V2G-enabled IoT devices may optimize energy distribution and ensure grid stability in the face of dynamic energy needs.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Cross Architecture Distillation for Face Recognition

arXiv

引用

arXiv 2023年

作者： Zhao, Weisong Zhu, Xiangyu He, Zhixiang Zhang, Xiao-Yu Lei, Zhen Institute of Information Engineering Chinese Academy of Sciences Beijing China CBSR&NLPR Institute of Automation Chinese Academy of Sciences Beijing China School of Cyber Security University of Chinese Academy of Sciences Beijing China School of Artificial Intelligence University of Chinese Academy of Sciences Beijing China China Telecom Corporation Ltd. Data & AI Technology Company China Centre for Artificial Intelligence and Robotics Hong Kong Institute of Science & Innovation Chinese Academy of Sciences Hong Kong

Transformers have emerged as the superior choice for face recognition tasks, but their insufficient platform acceleration hinders their application on mobile devices. In contrast, Convolutional Neural Networks (CNNs) capitalize on hardware-compatible acceleration libraries. Consequently, it has become indispensable to preserve the distillation efficacy when transferring knowledge from a Transformer-based teacher model to a CNN-based student model, known as Cross-Architecture Knowledge Distillation (CAKD). Despite its potential, the deployment of CAKD in face recognition encounters two challenges: 1) the teacher and student share disparate spatial information for each pixel, obstructing the alignment of feature space, and 2) the teacher network is not trained in the role of a teacher, lacking proficiency in handling distillation-specific knowledge. To surmount these two constraints, 1) we first introduce a Unified Receptive Fields Mapping module (URFM) that maps pixel features of the teacher and student into local features with unified receptive fields, thereby synchronizing the pixel-wise spatial information of teacher and student. Subsequently, 2) we develop an Adaptable Prompting Teacher network (APT) that integrates prompts into the teacher, enabling it to manage distillation-specific knowledge while preserving the model’s discriminative capacity. Extensive experiments on popular face benchmarks and two large-scale verification sets demonstrate the superiority of our method. Copyright © 2023, The Authors. All rights reserved.

关键词： Distillation

来源：评论

学校读者我要写书评

暂无评论

Connecting the Dots in Trustworthy artificial intelligence: From AI Principles, Ethics, and Key Requirements to Responsible AI Systems and Regulation

arXiv

引用

arXiv 2023年

作者： Díaz-Rodríguez, Natalia Ser, Javier Del Coeckelbergh, Mark de Prado, Marcos López Herrera-Viedma, Enrique Herrera, Francisco Department of Computer Science and Artificial Intelligence DaSCI Andalusian Institute in Data Science and Computational Intelligence University of Granada Granada18071 Spain Derio48160 Spain Bilbao48013 Spain Department of Philosophy University of Vienna Vienna1010 Austria School of Engineering Cornell University IthacaNY14850 United States ADIA Lab Al Maryah Island Abu Dhabi United Arab Emirates Department of Mathematics Khalifa University of Science and Technology Abu Dhabi United Arab Emirates

MSC Codes 68T01Trustworthy artificial intelligence (AI) is based on seven technical requirements sustained over three main pillars that should be met throughout the system’s entire life cycle: it should be (1) lawful, (2) ethical, and (3) robust, both from a technical and a social perspective. However, attaining truly trustworthy AI concerns a wider vision that comprises the trustworthiness of all processes and actors that are part of the system’s life cycle, and considers previous aspects from different lenses. A more holistic vision contemplates four essential axes: the global principles for ethical use and development of AI-based systems, a philosophical take on AI ethics, a risk-based approach to AI regulation, and the mentioned pillars and requirements. The seven requirements (human agency and oversight;robustness and safety;privacy and data governance;transparency;diversity, non-discrimination and fairness;societal and environmental wellbeing;and accountability) are analyzed from a triple perspective: What each requirement for trustworthy AI is, Why it is needed, and How each requirement can be implemented in practice. On the other hand, a practical approach to implement trustworthy AI systems allows defining the concept of responsibility of AI-based systems facing the law, through a given auditing process. Therefore, a responsible AI system is the resulting notion we introduce in this work, and a concept of utmost necessity that can be realized through auditing processes, subject to the challenges posed by the use of regulatory sandboxes. Our multidisciplinary vision of trustworthy AI culminates in a debate on the diverging views published lately about the future of AI. Our reflections in this matter conclude that regulation is a key for reaching a consensus among these views, and that trustworthy and responsible AI systems will be crucial for the present and future of our society. © 2023, CC BY-NC-ND.

关键词： artificial intelligence

来源：评论

学校读者我要写书评

暂无评论

Adaptive Cleaning in Manufacturing: A Decision Tree Model for Efficient Factory Sanitation

Adaptive Cleaning in Manufacturing: A Decision Tree Model fo...

引用

Innovative Practices in technology and Management (ICIPTM), International Conference on

作者： M. Arulprakash Ramakrishnan Raman Atul A Gokhale K. Saravanan Ishwarya M.V S. Sujatha Department of Computing Technologies SRM Institute of Science and Technology Kattankulathur Chenagalpattu Chennai Tamil Nadu India Symbiosis Institute of Business Management Pune & Symbiosis International (Deemed University) Pune Maharashtra India Symbiosis School of Culinary Arts Symbiosis International (Deemed University) Pune Maharashtra India Department of Information Technology R.M.D. Engineering College Chennai Tamil Nadu India Artificial Intelligence and Data Science Department Agni College of Technology Chennai Tamil Nadu India Department of Biomedical Engineering Saveetha School of Engineering Saveetha Institute of Medical and Technical Sciences Saveetha University Chennai Tamil Nadu India

ISBN: (数字)9798350307757

ISBN: (纸本)9798350307764

Industrial cleanliness is crucial in production settings, and this system offers a novel way to use Internet of Things (IoT) technology to strengthen this practice. The suggested solution uses a Decision Tree (DT) model and real-time data received from IoT sensors placed strategically around the facility to enable adaptive enhancements to cleaning processes. This system of sensors keeps tabs on a number of variables, including manufacturing schedules, environmental factors, and circulation. In real-time, the DT model may use this data to optimize cleaning locations, durations, and procedures, cutting down on wasteful downtime and resource usage without compromising cleanliness. When contrasted with conventional static cleaning methods, the adaptive cleaning system shows considerable gains in both sanitation efficiency and resource occupancy. Improving cleaning using IoT technology makes it more responsive and smart, which in turn makes production more efficient and less detrimental to the environment. It explains how the IoT and DT models worked together to make manufacturing processes more adaptable, under the lens of industrial hygiene.

关键词： Adaptation models Schedules Adaptive systems Cleaning Real-time systems Production facilities Environmental factors

来源：评论

学校读者我要写书评

暂无评论

BRAU-Net++: U-Shaped Hybrid CNN-Transformer Network for Medical Image Segmentation

arXiv

引用

arXiv 2024年

作者： Lan, Libin Cai, Pengzhou Jiang, Lu Liu, Xiaojuan Li, Yongmei Zhang, Yudong College of Computer Science and Engineering Chongqing University of Technology Chongqing400054 China College of Computer Science Chongqing University Chongqing400044 China College of Artificial Intelligence Chongqing University of Technology Chongqing401135 China Department of Radiology The First Affiliated Hospital of Chongqing Medical University Chongqing400016 China Department of Information Technology Faculty of Computing and Information Technology King Abdulaziz University Jeddah21589 Saudi Arabia School of Computer Science and Engineering Southeast University Jiangsu Nanjing210096 China School of Computing and Mathematical Sciences University of Leicester LeicesterLE1 7RH United Kingdom

Accurate medical image segmentation plays an essential role in clinical quantification, disease diagnosis, treatment planning and many other applications. Both convolution-based and transformer-based u-shaped architectures have made significant success in various medical image segmentation tasks. The former can efficiently learn local information of images while requiring much more image-specific inductive biases inherent to convolution operation. The latter can effectively capture long-range dependency at different feature scales using self-attention, whereas it typically encounters the challenges of quadratic compute and memory requirements with sequence length increasing. To address this problem, through integrating the merits of these two paradigms in a well-designed u-shaped architecture, we propose a hybrid yet effective CNN-Transformer network, named BRAUNet++, for an accurate medical image segmentation task. Specifically, BRAU-Net++ uses bi-level routing attention as the core building block to design our u-shaped encoder-decoder structure, in which both encoder and decoder are hierarchically constructed, so as to learn global semantic information while reducing computational complexity. Furthermore, this network restructures skip connection by incorporating channel-spatial attention which adopts convolution operations, aiming to minimize local spatial information loss and amplify global dimension-interaction of multi-scale features. Extensive experiments on three diverse imaging modalities datasets demonstrate that our proposed approach outperforms other state-of-the-art methods including its baseline: BRAU-Net under almost all evaluation metrics, which reveals the generality and robustness of our approach for multi-modal medical image segmentation tasks. The code and models are publicly available on GitHub. Copyright © 2024, The Authors. All rights reserved.

关键词： Convolutional neural networks

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

Domain-Specific Versus General-Purpose Word Representations in Sentiment Analysis for Deep Learning Models 7th

Domain-Specific Versus General-Purpose Word Representations ...

引用

7th International Conference on Frontiers of Intelligent Computing: Theory and Applications, FICTA 2018

作者： Hung, Bui Thanh Data Analytics & Artificial Intelligence Laboratory Engineering---Technology Faculty Thu Dau Mot University 6 Tran Van On Phu Hoa Thu Dau MotBinh Duong Viet Nam

ISBN: (纸本)9789813291850

Sentiment analysis, also known as opinion mining or emotion artificial intelligence, has been gradually becoming a hot-trend topic in the recent years due to its wide applications and also the challenges in handling the dataset efficiently. With the popular deep learning-based approach, using vectors of word as an input for the models is a good way for the machine to learn the relation between words in a document. In this paper, vector representations of word are created based on two different ways: learning from its own internal dataset, or deriving from a famous pre-trained library known as fastText. The two representations will then be used in documents (we classify sentiment on document level) to generate input to 4 deep learning models: Long-Short Term Memory Network (LSTM), Bi-directional Long-Short Term Memory Network (BiLSTM), Convolutional Neural Network (CNN) and Convolutional Neural Network combined with Long-Short Term Memory Network (CNN-LSTM). We do experiment in Vietnamese review dataset and explorer words in domain-specific and general–purpose before turning into vectors in two different ways: tokenized and not tokenized. The result shows the best result of each of the approaches in four deep learning models: LSTM, Bi-LSTM, CNN and CNN-LSTM. © 2020, Springer Nature Singapore Pte Ltd.

关键词： Sentiment 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：