检索结果-内蒙古大学图书馆

FetchAid: Making Parcel Lockers More Accessible to Blind and Low Vision People With Deep-learning Enhanced Touchscreen Guidance, Error-Recovery Mechanism, and AR-based Search Support

arXiv

引用

arXiv 2024年

作者： Guan, Zhitong Xiong, Zeyu Fan, Mingming Computational Media and Arts Thrust The Hong Kong University of Science and Technology Guangzhou China Division of Integrative Systems and Design Department of Computer Science and Engineering The Hong Kong University of Science and Technology Hong Kong

Parcel lockers have become an increasingly prevalent last-mile delivery method. Yet, a recent study revealed its accessibility challenges to blind and low-vision people (BLV). Informed by the study, we designed FetchAid, a standalone intelligent mobile app assisting BLV in using a parcel locker in real-time by integrating computer vision and augmented reality (AR) technologies. FetchAid first uses a deep network to detect the user's fingertip and relevant buttons on the touch screen of the parcel locker to guide the user to reveal and scan the QR code to open the target compartment door and then guide the user to reach the door safely with AR-based context-aware audio feedback. Moreover, FetchAid provides an error-recovery mechanism and real-time feedback to keep the user on track. We show that FetchAid substantially improved task accomplishment and efficiency, and reduced frustration and overall effort in a study with 12 BLV participants, regardless of their vision conditions and previous experience. © 2024, CC BY-NC-ND.

关键词： Augmented reality

来源：评论

学校读者我要写书评

暂无评论

Hessian Fly Infestation Assessment with Stress Indicator based on Multispectral Imaging and Machine Learning Techniques

Hessian Fly Infestation Assessment with Stress Indicator bas...

引用

2024 American Society of Agricultural and Biological Engineers Annual International Meeting, ASABE 2024

作者： Kwaku, Opoku-Ware Steven, Odubiyi Sanford, Eigenbrode Li, Liujun Department of Soil and Water Systems University of Idaho 875 Perimeter Drive MS 2060 MoscowID83844-2060 United States Department of Entomology Plant Pathology & Nematology 875 Perimeter Drive MS 2060 MoscowID83844-2060 United States Department of Mechanical Engineering University of Idaho MoscowID United States Department of Computer Science University of Idaho MoscowID United States

ISBN: (纸本)9798331302214

Remote sensing technologies and vegetation indices play a crucial role in the early detection and monitoring of biotic stresses, such as Hessian fly infestation in wheat plants. By analyzing and comparing the Normalized Difference Vegetation Index (NDVI) and Normalized Difference Red Edge Index (NDRE) in Hessian fly infestation treatment and control plants, we aimed to evaluate these indices as indicators of plant health. The study involved the use of multispectral images to calculate NDVI and NDRE in selected regions of interest. The results showed that on the 14th day after infestation, NDVI for infected plants ranged from 0.2947 to 0.4577, while NDVI for non-infected plants ranged from 0.3464 to 0.4689. On the 21st day, both NDVI and NDRE declined, with a more significant drop in infected plants. A K-nearest neighbor (KNN) classifier machine learning model was developed to categorize plant health status based on NDVI and NDRE values, achieving 80% and 70% accuracy, respectively. A Support Vector Machine (SVM) model with polynomial and radial basis function (RBF) kernels was also applied, achieving accuracies up to 90%. The study demonstrated that the pronounced reduction in NDVI and NDRE in infected plants indicates deteriorating canopy greenness and chlorophyll content, suggesting worsening plant health. The accuracy of the KNN and SVM classifiers highlight the potential of combining remote sensing with machine learning for monitoring and assessing plant diseases. Overall, the developed multispectral imaging and machine learning models provides an effective means of identifying the progression of Hessian fly infestation over time and offers a framework for assessing plant health and classify disease occurrence/severity, enabling prompt responses decision making for precision farming. © 2024 ASABE Annual International Meeting. All rights reserved.

关键词： Nearest neighbor search

来源：评论

学校读者我要写书评

暂无评论

Multimodal Classification and Out-of-distribution Detection for Multimodal Intent Understanding

arXiv

引用

arXiv 2024年

作者： Zhang, Hanlei Zhou, Qianrui Xu, Hua Su, Jianhua Evans, Roberto Gao, Kai State Key Laboratory of Intelligent Technology and Systems Department of Computer Science and Technology Tsinghua University Beijing100084 China School of Information Science and Engineering Hebei University of Science and Technology Shijiazhuang050018 China

Multimodal intent understanding is a significant research area that requires effectively leveraging multiple modalities to analyze human language. Existing methods face two main challenges in this domain. Firstly, they have limitations in capturing nuanced and high-level semantics underlying complex in-distribution (ID) multimodal intents. Secondly, they exhibit poor generalization when confronted with unseen out-of-distribution (OOD) data in real-world scenarios. To address these issues, we propose a novel method for both ID classification and OOD detection (MIntOOD). We first introduce a weighted feature fusion network that models multimodal representations effectively. This network dynamically learns the importance of each modality, adapting to multimodal contexts. To develop discriminative representations that are conducive to both tasks, we synthesize pseudo-OOD data from convex combinations of ID data and engage in multimodal representation learning from both coarse-grained and fine-grained perspectives. The coarse-grained perspective focuses on distinguishing between ID and OOD binary classes, while the fine-grained perspective enhances the understanding of ID data by incorporating binary confidence scores. These scores help to gauge the difficulty of each sample, improving the classification of different ID classes. Additionally, the fine-grained perspective captures instance-level interactions between ID and OOD samples, promoting proximity among similar instances and separation from dissimilar ones. We establish baselines for three multimodal intent datasets and build an OOD benchmark. Extensive experiments on these datasets demonstrate that our method significantly improves OOD detection performance with a 3-10% increase in AUROC scores while achieving new state-of-the-art results in ID classification. 1 Copyright © 2024, The Authors. All rights reserved.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Correction to: Convolutional neural network and ensemble machine learning model for optimizing performance of emotion recognition in wild (Multimedia Tools and Applications, (2023), 10.1007/s11042-023-17744-1)

引用

Multimedia Tools and Applications 2024年 1-1页

作者： Alturki, Nazik Umer, Muhammad Alshardan, Amal Saidani, Oumaima Abate, Andrea F. Ashraf, Imran Department of Information Systems College of Computer and Information Sciences Princess Nourah Bint Abdulrahman University P.O. Box 84428 Riyadh11671 Saudi Arabia Department of Computer Science The Islamia University of Bahawalpur Bahawalpur63100 Pakistan Department of Computer Science University of Salerno Fisciano84084 Italy Information and Communication Engineering Yeungnam University Gyeongsan38541 Korea Republic of

The authors Muhammad Umer and Oumaima Saidani have two affiliations in the original publication of this article which was incorrect. The original article has been corrected. © 2024, The Author(s), under exclusive...

关键词：

来源：评论

学校读者我要写书评

暂无评论

Solving the N-Queens and Golomb Ruler Problems Using DQN and an Approximation of the Convergence 28th

Solving the N-Queens and Golomb Ruler Problems Using DQN a...

引用

28th International Conference on Neural Information Processing, ICONIP 2021

作者： Prudhvi Raj, Patnala Saha, Snehanshu Srinivasa, Gowri PES Center for Pattern Recognition and Department of Computer Science and Engineering PES University Bengaluru India Department of Computer Science and Information Systems and APPCAIR BITS Pilani K K Birla Goa Campus Goa India

ISBN: (纸本)9783030923099

We build on the Deep Q-Learning Network (DQN) to solve the N-Queens problem to propose a solution to the Golomb Ruler problem, a popular example of a one dimensional constraint satisfaction problem. A comparison of the DQN approach with standard solution approaches to solve constraint satisfaction problems, such as backtracking and branch-and-bound, demonstrates the efficacy of the DQN approach, with significant computational savings as the order of the problem increases. The convergence behavior of the DQN model has been approximated using Locally Weighted Regression and Cybenko Approximation, demonstrating an improvement in the performance of the DQN with episodes, regardless of the order of the problem. © 2021, Springer Nature Switzerland AG.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

A Machine Learning Method for Detecting Sarcasm in Negative Sentiment Tweets using Hyperboles

A Machine Learning Method for Detecting Sarcasm in Negative ...

引用

2022 IEEE North Karnataka Subsection Flagship International Conference, NKCon 2022

作者： Preetha, V. Sathishkumar, P. Janani, M. Manohari, B. Ida, S. Jhansi Suganyadevi, S. Sri Ramakrishna Institute of Technology Department of CSE Coimbatore India Vel Tech Multi Tech Dr.Rangarajan Dr.Sakunthala Engineering College Avadi Department of CSE Chennai India KPR Institute of Engineering and Technology Department of CSE Coimbatore India Sri Krishna College of Engineering and Technology Department of M.Tech Computer Science and Engineering Coimbatore India R.M.K Engineering Colleg Department of Computer Science and Business Systems Kavaraipettai India KPR Institute of Engineering and Technology Department of ECE Coimbatore India

The use of sarcasm as an expression is a common literary device that involves someone purposefully expressing the opposite of what is being meant. The accurate identification of sarcasm in a text may boost other natural language processing activities, particularly ones involving sentiment analysis, and make it easier to comprehend the genuine intents of the speakers in the text. It is difficult to determine if someone is being sarcastic only based on their written words since sarcasm is a kind of implicitly expressing one's feelings and sarcastic speakers purposefully confound their audiences. Existing methods based on machine learning and deep learning obtained an insufficient level of performance when processing sarcastic material that included complicated expressions or required a certain set of prior knowledge to comprehend. The purpose of this research is to examine how sarcasm may be detected in tweets that have a negative mood and include hyperboles. For the purpose of sarcasm identification, the hashtags #Chinesevirus, #Kungflu, #COVID19, #Hantavirus, and #Coronavirus were included in the collection of 6,600 pre-processed negative sentiment tweets. The three well-known machine learning algorithms known as Support Vector Machine, Random Forest, and Random Forest with Bagging were utilised in order to conduct an analysis of the five hyperbole characteristics known as interjection, intensifier, capital letter, punctuation mark, and elongated word. The suggested model with the extended term attained an accuracy of 78.74% and an F-score of 71%, respectively, despite the inclusion of hyperbolic words in the tweets included in an unbiased dataset. It was discovered that intensifier was the exaggeration with the highest level of significance (p. 0001). The results of the experiments and analyses carried out for this study led the researchers to the conclusion that hyperboles may be found in unbiased datasets, which helps improve sarcasm identification as well. © 202

关键词： Sentiment analysis

来源：评论

学校读者我要写书评

暂无评论

A Comparative Evaluation of Ensemble Classifiers for Malicious Webpage Detection 18

A Comparative Evaluation of Ensemble Classifiers for Malicio...

引用

18th International Learning and Technology Conference, L and T 2021

作者： Subasi, Abdulhamit Balfaqih, Mohammed Balfagih, Zain Alfawwaz, Khaled Computer Science Department Effat University Jeddah Saudi Arabia Department of Computer and Network Engineering University of Jeddah Jeddah Saudi Arabia Department of Information Systems King Abdulaziz University Saudi Arabia

Malicious webpage is developed or manipulated to be used as attack tool where it is considered as one of the main reasons of Internet criminal activities. Thus, it is essential to detect such webpages and prevent end users form accessing it. The conventional malicious webpages detection techniques are based on searching through a blacklist that contains a list of webpages classified as malicious from the perspective of users. However, these techniques have high false-negative rates especially with aforesaid sophisticated attacks due to technical and computational limitations. Hence, machine learning techniques have been employed to classify webpages by systemically analyzing set of features that reflect the characteristics of a malicious webpage. This paper compares the prediction accuracy of several machine learning classification algorithms and ensemble techniques. A data set of 5000 instances of URLs, with 189 different features are used in the comparative study. The results show that the most accurate classification technique in MultiBoost and Adaboost is Support Vector Machine (SVM), while K-Nearest Neighbor (k-NN) technique in bagging and random subspace. © 2021 The Authors. Published by Elsevier B.V.

关键词： Face recognition

来源：评论

学校读者我要写书评

暂无评论

Medical Feature Selection Approach Based on Generalized Normal Distribution Algorithm

引用

computers, Materials & Continua 2021年第12期69卷 2883-2901页

作者： Mohamed Abdel-Basset Reda Mohamed Ripon K.Chakrabortty Michael J.Ryan Yunyoung Nam Mohamed Abouhawwash Faculty of Computers and Informatics Zagazig UniversityZagazig44519Egypt Capability Systems Centre School of Engineering and ITUNSWCanberraAustralia Department of Computer Science and Engineering Soonchunhyang UniversityAsan31538Korea Department of Mathematics Faculty of ScienceMansoura UniversityMansoura35516Egypt Department of Computational Mathematics Scienceand Engineering(CMSE)Michigan State UniversityEast Lansing48824MIUSA

This paper proposes a new pre-processing technique to separate the most effective features from those that might deteriorate the performance of the machine learning classifiers in terms of computational costs and classification accuracy because of their irrelevance,redundancy,or less information;this pre-processing process is often known as feature *** technique is based on adopting a new optimization algorithm known as generalized normal distribution optimization(GNDO)supported by the conversion of the normal distribution to a binary one using the arctangent transfer function to convert the continuous values into binary ***,a novel restarting strategy(RS)is proposed to preserve the diversity among the solutions within the population by identifying the solutions that exceed a specific distance from the best-so-far and replace them with the others created using an effective updating *** strategy is integrated with GNDO to propose another binary variant having a high ability to preserve the diversity of the solutions for avoiding becoming stuck in local minima and accelerating convergence,namely improved GNDO(IGNDO).The proposed GNDO and IGNDO algorithms are extensively compared with seven state-of-the-art algorithms to verify their performance on thirteen medical instances taken from the UCI *** is shown to be superior in terms of fitness value and classification accuracy and competitive with the others in terms of the selected *** the principal goal in solving the FS problem is to find the appropriate subset of features that maximize classification accuracy,IGNDO is considered the best.

关键词： Generalized normal distribution optimization feature selection transfer function novel restarting strategy UCI repository

来源：评论

学校读者我要写书评

暂无评论

Medical image segmentation approach based on hybrid adaptive differential evolution and crayfish optimizer

引用

computers in Biology and Medicine 2024年 180卷 109011-109011页

作者： Mostafa, Reham R. Khedr, Ahmed M. Aghbari, Zaher AL Afyouni, Imad Kamel, Ibrahim Ahmed, Naveed University of Sharjah Sharjah27272 United Arab Emirates Information Systems Department Faculty of Computers and Information Sciences Mansoura University Mansoura35516 Egypt Computer Science Department University of Sharjah Sharjah27272 United Arab Emirates Electrical & Computer Engineering Department University of Sharjah Sharjah27272 United Arab Emirates

Image segmentation plays a pivotal role in medical image analysis, particularly for accurately isolating tumors and lesions. Effective segmentation improves diagnostic precision and facilitates quantitative analysis, which is vital for medical professionals. However, traditional segmentation methods often struggle with multilevel thresholding due to the associated computational complexity. Therefore, determining the optimal threshold set is an NP-hard problem, highlighting the pressing need for efficient optimization strategies to overcome these challenges. This paper introduces a multi-threshold image segmentation (MTIS) method that integrates a hybrid approach combining Differential Evolution (DE) and the Crayfish Optimization Algorithm (COA), known as HADECO. Utilizing two-dimensional (2D) Kapur's entropy and a 2D histogram, this method aims to enhance the efficiency and accuracy of subsequent image analysis and diagnosis. HADECO is a hybrid algorithm that combines DE and COA by exchanging information based on predefined rules, leveraging the strengths of both for superior optimization results. It employs Latin Hypercube Sampling (LHS) to generate a high-quality initial population. HADECO introduces an improved DE algorithm (IDE) with adaptive and dynamic adjustments to key DE parameters and new mutation strategies to enhance its search capability. In addition, it incorporates an adaptive COA (ACOA) with dynamic adjustments to the switching probability parameter, effectively balancing exploration and exploitation. To evaluate the effectiveness of HADECO, its performance is initially assessed using CEC’22 benchmark functions. HADECO is evaluated against several contemporary algorithms using the Wilcoxon signed rank test (WSRT) and the Friedman test (FT) to integrate the results. The findings highlight HADECO's superior optimization abilities, demonstrated by its lowest average Friedman ranking of 1.08. Furthermore, the HADECO-based MTIS method is evaluated using M

关键词： Image segmentation

来源：评论

学校读者我要写书评

暂无评论

Towards Consistent and Explainable Motion Prediction using Heterogeneous Graph Attention

Towards Consistent and Explainable Motion Prediction using H...

引用

IEEE Symposium on Intelligent Vehicle

作者： Tobias Demmler Andreas Tamke Thao Dang Karsten Haug Lars Mikelsons Robert Bosch GmbH Stuttgart Germany Department of Computer Science and Engineering Institute for Intelligent Systems Esslingen University Germany Leads the Chair of Mechatronics University of Augsburg Germany

ISBN: (数字)9798350348811

ISBN: (纸本)9798350348828

In autonomous driving, accurately interpreting the movements of other road users and leveraging this knowledge to forecast future trajectories is crucial. This is typically achieved through the integration of map data and tracked trajectories of various agents. Numerous methodologies combine this information into a singular embedding for each agent, which is then utilized to predict future behavior. However, these approaches have a notable drawback in that they may lose exact location information during the encoding process. The encoding still includes general map information. However, the generation of valid and consistent trajectories is not guaranteed. This can cause the predicted trajectories to stray from the actual lanes. This paper introduces a new refinement module designed to project the predicted trajectories back onto the actual map, rectifying these discrepancies and leading towards more consistent predictions. This versatile module can be readily incorporated into a wide range of architectures. Additionally, we propose a novel scene encoder that handles all relations between agents and their environment in a single unified heterogeneous graph attention network. By analyzing the attention values on the different edges in this graph, we can gain unique insights into the neural network’s inner workings leading towards a more explainable prediction.

关键词： Measurement Tracking Roads Encoding Trajectory Autonomous vehicles

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：