检索结果-内蒙古大学图书馆

A Universal Spatiotemporal Feature Fusion Model for Gait Recognition

学校读者我要写书评

暂无评论

SSRN

SSRN 2023年

作者： Xu, Wei Li, Kang School of Biomedical Engineering Division of Life Sciences and Medicine University of Science and Technology of China Hefei China Suzhou Institute for Advanced Research University of Science and Technology of China Suzhou China West China Biomedical Big Data Center West China Hospital Sichuan University Chengdu China Shanghai Artificial Intelligence Laboratory Shanghai China

Gait recognition has emerged as a crucial biometric technology that can identify the walking pattern of an individual from a distance and without their cooperation. In recent years, this field has garnered increased attention in research. Most contemporary gait recognition approaches rely on extracting gait feature descriptors from global appearance and human local regions, while others incorporate temporal feature information to enhance recognition accuracy. Despite the advancements in gait recognition, recent frameworks have been utilizing increasingly complex network structures to extract the necessary features. As a result, these models can be challenging to apply in real-world scenarios, necessitating the development of a more practical and efficient network architecture. In this paper, we present a novel Universal Spatiotemporal Feature Fusion baseline that enhances spatiotemporal feature fusion for gait recognition. Our approach achieves state-of-the-art performance on multiple datasets and strives to offer a feasible and reliable solution for gait recognition, catering to various practical applications. © 2023, The Authors. All rights reserved.

关键词： Complex networks

ViGT: Proposal-free Video Grounding with Learnable Token in Transformer

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Li, Kun Guo, Dan Wang, Meng School of Computer Science and Information Engineering Hefei University of Technology Hefei230601 China Key Laboratory of Knowledge Engineering with Big Data Ministry of Education Hefei230601 China Intelligent Interconnected Systems Laboratory of Anhui Province Hefei230601 China Institute of Artificial Intelligence Hefei Comprehensive National Science Center Hefei230088 China

The video grounding (VG) task aims to locate the queried action or event in an untrimmed video based on rich linguistic descriptions. Existing proposal-free methods are trapped in complex interaction between video and query, overemphasizing cross-modal feature fusion and feature correlation for VG. In this paper, we propose a novel boundary regression paradigm that performs regression token learning in a transformer. Particularly, we present a simple but effective proposal-free framework, namely Video Grounding Transformer (ViGT), which predicts the temporal boundary using a learnable regression token rather than multi-modal or cross-modal features. In ViGT, the benefits of a learnable token are manifested as follows. (1) The token is unrelated to the video or the query and avoids data bias toward the original video and query. (2) The token simultaneously performs global context aggregation from video and query features. First, we employed a sharing feature encoder to project both video and query into a joint feature space before performing cross-modal co-attention (i.e., video-to-query attention and query-to-video attention) to highlight discriminative features in each modality. Furthermore, we concatenated a learnable regression token [REG] with the video and query features as the input of a vision-language transformer. Finally, we utilized the token [REG] to predict the target moment and visual features to constrain the foreground and background probabilities at each timestamp. The proposed ViGT performed well on three public datasets: ANet Captions, TACoS and YouCookII. Extensive ablation studies and qualitative analysis further validated the interpretability of ViGT. Copyright © 2023, The Authors. All rights reserved.

关键词： Regression analysis

Communication Induced Checkpointing based Fault Tolerance Mechanism – A Review and CIAC-FTM Framework in IoT Environment

学校读者我要写书评

暂无评论

Communication Induced Checkpointing based Fault Tolerance Me...

International Conference on Computing, Communication, and Intelligent Systems (ICCCIS)

作者： A Sowjanya Lakshmi Ch Vani Priya Gaurav Gupta Information Science and Engineering Sir. M. Visvesvaraya Institute of Technology Bengaluru India Yogananda School of Artificial Intelligence Computer and Data Sciences Shoolini University Solan Himachal Pradesh India

ISBN: (纸本)9781665462013

High performance computations in IoTsystems processes huge data, suffers from various types of faults due to hardware or software faults, malicious attacks, network congestion, missing deadlines and server overloads, and Fault Tolerance in such systems is mandatory to maintain the performance of long running *** Induced Adaptive Checkpointing Fault Tolerance Mechanism (CIAC-FTM) used to address such Faults. The research focus on Software Faults - Transient Faults at application level. This CIAC-FTM places checkpoints at required nodes depending on the type of fault detected, in turn reduces checkpoints and storage overheads. The collaboration of Sensors, huge data, Machine to Machine (M2M) communication,and IoT gives us a promising formula for near perfect system. Current technologies are able to work hand-in-hand to create a bigger system. IoT ensures a more connected world. The data procured by the sensor nodes are pushed into a network and processing units, processed data used to take decision to act on actuators as designed. Fault-free data at every transaction is at most priority and so Communication Induced Checkpoints (CIC)areintroduced to ensure that the fault-free task restarts and hence CIAC-FTM ensure that the transient faults are rectified.

关键词： Checkpointing Fault tolerance Adaptive systems Fault tolerant systems Software Sensor systems Sensors

Efficient hybrid modeling and sorption model discovery for non-linear advection-diffusion-sorption systems: A systematic scientific machine learning approach

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Santana, Vinicius Viena Costa, Erbet de Menezes Rebello, Carine Ribeiro, Ana Mafalda Rackauckas, Chris Nogueira, Idelfonso B.R. Department of Chemical Engineering Norwegian University of Science and Technology Norway JuliaLab Computer Science and Artificial Intelligence Laboratory Massachusetts Institute of Technology United States LSRE - LCM Laboratory of Separation and Reaction Engineering – Laboratory of Catalysis and Materials Faculty of Engineering University of Porto Rua Dr. Roberto Frias Porto4200-465 Portugal ALiCE Associate Laboratory in Chemical Engineering Faculty of Engineering University of Porto Rua Dr. Roberto Frias Porto4200-465 Portugal

This study presents a systematic machine learning approach for creating efficient hybrid models and discovering sorption uptake models in non-linear advection-diffusion-sorption systems. It demonstrates an effective method to train these complex systems using gradient-based optimizers, adjoint sensitivity analysis, and JIT-compiled vector Jacobian products, combined with spatial discretization and adaptive integrators. Sparse and symbolic regression were employed to identify missing functions in the artificial neural network. The robustness of the proposed method was tested on an in-silico data set of noisy breakthrough curve observations of fixed-bed adsorption, resulting in a well-fitted hybrid model. The study successfully reconstructed sorption uptake kinetics using sparse and symbolic regression, and accurately predicted breakthrough curves using identified polynomials, highlighting the potential of the proposed framework for discovering sorption kinetic law structures. © 2023, CC BY.

关键词： Sorption

Predicting of Uncertainty Facial Expressions in Children with Autism Spectrum Disorder Based on an Improved Fuzzy Convolutional Neural Network Model

学校读者我要写书评

暂无评论

SSRN

SSRN 2024年

作者： Abd-Alhalem, Samia M. Ali, Hesham A. El-Shafai, Walid Algarni, Abeer D. Soliman, Naglaa F. El-Samie, Fathi E. Faculty of Artificial Intelligence Delta University for Science and Technology Gamasa35712 Egypt Computers and Control Systems Engineering Department Faculty of Engineering Mansoura University Mansoura35516 Egypt Security Engineering Lab Computer Science Department Prince Sultan University Riyadh11586 Saudi Arabia Department Electronics and Electrical Communications Engineering Faculty of Electronic Engineering Menoufia University Menouf32952 Egypt Department of Information Technology College of Computer and Information Sciences Princess Nourah Bint Abdulrahman University P.O. Box 84428 Riyadh11671 Saudi Arabia Zagazig University Egypt Menoufia University Egypt

Facial expression recognition plays a pivotal role in comprehending emotions and social interactions, particularly in individuals grappling with autism spectrum disorder (ASD), who often encounter challenges in non-verbal communication. For some with ASD, even subtle shifts in their facial expressions can prove elusive, rendering it arduous to categorize their emotional states accurately. Consequently, the analysis and processing of facial expression data hold the potential to facilitate early autism diagnosis, thereby enabling timely therapeutic interventions and alleviating the economic and psychological burdens experienced by patients. In this study, we introduce an enhanced approach that leverages a fuzzy convolutional neural network (CNN) and an Improved Binary Genetic Algorithm (IBGA) for autism diagnosis based on facial expression data. Our model confronts the intricate task of recognizing facial images by selecting salient features employing the IBGA method. Moreover, it employs fuzzy weights to ascertain the probability of similarity within the classifier layer. Our proposed model has yielded impressive outcomes, boasting an average accuracy of 0.9405 for ASD detection and 0.9365 for the recognition of six distinct facial expressions. These results hold profound implications for the enhancement of emotion recognition systems and the refinement of interventions aimed at bolstering social interactions for individuals grappling with ASD. © 2024, The Authors. All rights reserved.

关键词： Emotion Recognition

Fake Profile Identification Using Machine Learning

学校读者我要写书评

暂无评论

Fake Profile Identification Using Machine Learning

IEEE International WIE Conference on Electrical and Computer engineering (WIECON-ECE)

作者： T. Sudhakar Bhuvana Chendrica Gogineni J. Vijaya School of Computer Science and Engineering VIT-AP University Vijayawada Andrapradesh School of Computer Science and Engineering Ira A. Fulton Schools of Engineering Arizona State University United States Department Data Science and Artificial Intelligence International Institute of Information Technology Naya Raipur Chhattisgarh

Online social networks have permeated our social lives in the current generation. These sites have allowed us to see our social lives differently than they did in the past. Nowadays we can connect with new friends and maintain relationships with them via social and personal activities become quite easy. Online Social Networks (OSN) are contributed in all areas such as Research in all domains, Job-related areas, technology oriented areas, Health care, and business-oriented areas, Information gathering and data collection, and so on. One of the biggest problems on these social media platforms is fake profiles. Impersonating to be someone else and causing harm and defamation to the real person or advertising or popularizing removed propaganda on someone’s name to get more benefit is the motto of such profile creators. There have been many studies regarding these fake accounts and how can they be mitigated. Many approaches such as graph-level activities or feature analysis have been taken into consideration to identify fake profiles. These methods are outdated when compared to arising issues of these days. In this paper, we proposed a technique using machine learning for fake profile detection which is efficient. . The benchmark data set is collected and mixed with manual data first furthermore; a data cleaning technique is used to present the data more feasibly. Then the preprocessed data is used for model building with sufficient information such as profile name, profile ID name, number of followers, and so on. We added Cross validation process where many training algorithms are implemented on the given data and are then tested on the same data. Based on the experiments the RF classifier performed better than the other classification methods. The Random Forest classifier is used to forecast the profile whether is fake or genuine in an efficient way.

关键词：

EDOLAB: An Open-Source Platform for Education and Experimentation with Evolutionary Dynamic Optimization Algorithms

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Peng, Mai Yazdani, Delaram She, Zeneng Yazdani, Danial Luo, Wenjian Li, Changhe Branke, Juergen Nguyen, Trung Thanh Gandomi, Amir H. Yang, Shengxiang Jin, Yaochu Yao, Xin School of Automation China University of Geosciences Hubei Key Laboratory of Advanced Control and Intelligent Automation for Complex Systems Engineering Research Center of Intelligent Technology for Geo-Exploration Ministry of Education Wuhan China Research Institute Faculty of Engineering and Technology Liverpool John Moores University United Kingdom School of Computer Science and Technology Harbin Institute of Technology China Guangdong Provincial Key Laboratory of Novel Security Intelligence Technologies School of Computer Science and Technology Harbin Institute of Technology Peng Cheng Laboratory China School of Artificial Intelligence Anhui University of Sciences & Technology China Information Systems Management and Analytics in Warwick Business School University of Warwick United Kingdom Faculty of Engineering & Information Technology University of Technology Sydney Australia Obuda University Hungary School of Computer Science and Informatics De Montfort University United Kingdom Department of Artificial Intelligence School of Engineering Westlake University China School of Computer Science University of Birmingham United Kingdom

Many real-world optimization problems exhibit dynamic characteristics, posing significant challenges for traditional optimization techniques. Evolutionary Dynamic Optimization Algorithms (EDOAs) are designed to address these challenges effectively. However, in existing literature, the reported results for a given EDOA can vary significantly. This inconsistency often arises because the source codes for many EDOAs, which are typically complex, have not been made publicly available, leading to error-prone re-implementations. To support researchers in conducting experiments and comparing their algorithms with various EDOAs, we have developed an open-source MATLAB platform called the Evolutionary Dynamic Optimization LABoratory (EDOLAB). This platform not only facilitates research but also includes an educational module designed for instructional purposes. The education module allows users to observe: a) a 2-dimensional problem space and its morphological changes following each environmental change, b) the behaviors of individuals over time, and c) how the EDOA responds to environmental changes and tracks the moving optimum. The current version of EDOLAB features 25 EDOAs and four fully parametric benchmark generators. © 2023, CC BY-NC-SA.

关键词： Optimization algorithms

IoT's Role in Personalized Dietary Support for Mental Health using RNN Model

学校读者我要写书评

暂无评论

IoT's Role in Personalized Dietary Support for Mental Health...

Electrical, Electronics, Information and Communication Technologies (ICEEICT), International Conference on

作者： S Subha N.Mohankumar Allanki Sanyasi Rao A. Ayub Khan Ishwarya M. V S. Murugan Department of English M Kumarasamy College of Engineering Karur Tamil Nadu India Symbiosis Institute of Technology (SIT) Symbiosis International (Deemed University) Nagpur Maharashtra India Department of Electronics and Communication Engineering Christu Jyothi Institute of Technology & Science Jangaon Telangana India Department of Electronics and Communication Engineering Sona College of Technology Salem Tamil Nadu India Artificial Intelligence and Data Science Department Agni College of Technology Thazhambur Tamil Nadu Department of Biomedical Engineering Saveetha School of Engineering Saveetha Institute of Medical and Technical Sciences Saveetha University Chennai Tamil Nadu India

ISBN: (数字)9798350369083

ISBN: (纸本)9798350369090

Rapid developments in Internet of Things (IoT) technology have spawned new ways of thinking about the very important field of mental health and wellbeing. This abstract explores "Personalized Dietary Support," a new paradigm powered by the Internet of Things that aims to improve mental health outcomes by providing individualized dietary counseling. Individuals may get real-time insights into the dynamic interaction between their food choices, lifestyle, and mental well-being using data from IoT devices, such as wearables, nutritional trackers, and physiological sensors. Incorporating nutrient intake, meal timing, and dietary behaviors favorable to enhanced mental health, machine learning algorithms evaluate this data to provide tailored dietary recommendations. The IoT ecosystem also encourages user engagement through gamification and constant progress monitoring, which helps to keep users committed to their new eating habits over time. Working with medical experts, patients may care for their mental health from every angle. Recognizing the important importance of diet on mental health, this abstract investigates the possibility of IoT-driven Personalized Dietary Support to equip people with implementable methods. data privacy and security are emphasized, as are patient autonomy and systemic approaches to healthcare.

关键词： data privacy Machine learning algorithms Mental health Medical services Real-time systems Timing Internet of Things Wearable devices Wearable sensors Monitoring

BioT5: Enriching Cross-modal Integration in Biology with Chemical Knowledge and Natural Language Associations

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Pei, Qizhi Zhang, Wei Zhu, Jinhua Wu, Kehan Gao, Kaiyuan Wu, Lijun Xia, Yingce Yan, Rui Gaoling School of Artificial Intelligence Renmin University of China China University of Science and Technology of China China Huazhong University of Science and Technology China Microsoft Research United States Engineering Research Center of Next-Generation Intelligent Search and Recommendation Ministry of Education China Beijing Key Laboratory of Big Data Management and Analysis Methods China

Recent advancements in biological research leverage the integration of molecules, proteins, and natural language to enhance drug discovery. However, current models exhibit several limitations, such as the generation of invalid molecular SMILES, underutilization of contextual information, and equal treatment of structured and unstructured knowledge. To address these issues, we propose BioT5, a comprehensive pretraining framework that enriches cross-modal integration in biology with chemical knowledge and natural language associations. BioT5 utilizes SELFIES for 100% robust molecular representations and extracts knowledge from the surrounding context of bio-entities in unstructured biological literature. Furthermore, BioT5 distinguishes between structured and unstructured knowledge, leading to more effective utilization of information. After fine-tuning, BioT5 shows superior performance across a wide range of tasks, demonstrating its strong capability of capturing underlying relations and properties of bio-entities. Our code is available at https://***/QizhiPei/BioT5. Copyright © 2023, The Authors. All rights reserved.

关键词： Integration