检索结果-内蒙古大学图书馆

12th International Conference on Learning Representations, ICLR 2024

作者： Kesen, Ilker Pedrotti, Andrea Dogan, Mustafa Cafagna, Michele Acikgoz, Emre Can Parcalabescu, Letitia Calixto, Iacer Frank, Anette Gatt, Albert Erdem, Aykut Erdem, Erkut Koç University KUIS AI Center Turkey Koç University Department of Computer Engineering Turkey University of Pisa Department of Computer Science Italy Institute of Information Science and Technologies Italian National Council of Research Italy Hacettepe University Department of Computer Engineering Turkey Aselsan Research Turkey University of Malta Institute of Linguistics and Language Technology Malta Heidelberg University Department of Computational Linguistics Germany Amsterdam UMC University of Amsterdam Department of Medical Informatics Netherlands Amsterdam Public Health Methodology & Mental Health Amsterdam Netherlands Utrecht University Department of Information and Computing Sciences Netherlands

With the ever-increasing popularity of pretrained Video-Language Models (VidLMs), there is a pressing need to develop robust evaluation methodologies that delve deeper into their visio-linguistic capabilities. To address this challenge, we present VILMA), a task-agnostic benchmark that places the assessment of fine-grained capabilities of these models on a firm footing. Task-based evaluations, while valuable, fail to capture the complexities and specific temporal aspects of moving images that VidLMs need to process. Through carefully curated counterfactuals, VILMA offers a controlled evaluation suite that sheds light on the true potential of these models, as well as their performance gaps compared to human-level understanding. VILMA also includes proficiency tests, which assess basic capabilities deemed essential to solving the main counterfactual tests. We show that current VidLMs' grounding abilities are no better than those of vision-language models which use static images. This is especially striking once the performance on proficiency tests is factored in. Our benchmark serves as a catalyst for future research on VidLMs, helping to highlight areas that still need to be explored. © 2024 12th International Conference on Learning Representations, ICLR 2024. All rights reserved.

关键词： Computational linguistics

来源：评论

学校读者我要写书评

暂无评论

Deep Learning for Violence Detection in Surveillance: The Role of Transfer Learning and Pre-Trained Models

Deep Learning for Violence Detection in Surveillance: The Ro...

引用

Information Technology (ACIT)

作者： Anum Ilyas Surayya Obaid Narmeen Zakaria Bawany Department of Computer Science and Software Engineering Center for Computing Research Jinah University for Women Karachi Pakistan

In recent years, the use of CCTV footage for proactive crime prevention has surged, particularly in public places like airports, train stations, and malls. However, the efficacy of these surveillance systems becomes questionable for their substantial reliance on human resources which may lead to erroneous or delayed responses. This research takes a comprehensive approach for violence detection which was previously limited to binary classification, by categorizing violence into four distinct classes: abuse, arson, assault, and fight. It employs transfer learning combined with computer vision techniques for violence detection and classification in video footages. The study compares the performance of four pre-trained neural networks, namely DenseNet121, VGG16, MobileNet, and Xception. The dataset created for this research is compiled from three datasets available on Kaggle. The results reveal that the Xception model performed comparatively better achieving the highest AUC score of 98.39%, while the VGG16 model attained the lowest AUC score of 96%. In addition to the AUC score, precision, recall, and fl-score are employed as performance metrics. Transfer learning with convolutional neural networks (CNN) significantly reduced computational requirements and time. Automating the detection and categorization of violent behavior through the employed approach has the potential to reduce the risk of fatalities and injuries in public areas. It also improves the speed and accuracy of threat detection, enabling swift preventive actions by authorities.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Reactor Optimization Benchmark by Reinforcement Learning

arXiv

引用

arXiv 2024年

作者： Schwarcz, Deborah Schneider, Nadav Oren, Gal Steinitz, Uri Soreq Nuclear Research Center Israel Scientific Computing Center Nuclear Research Center Negev Israel Department of Electrical & Computer Engineering Ben-Gurion University Israel Department of Computer Science Technion - Israel Institute of Technology Israel

Neutronic calculations for reactors are a daunting task when using Monte Carlo (MC) methods. As high-performance computing has advanced, the simulation of a reactor is nowadays more readily done, but design and optimization with multiple parameters is still a computational challenge. MC transport simulations, coupled with machine learning techniques, offer promising avenues for enhancing the efficiency and effectiveness of nuclear reactor optimization. This paper introduces a novel benchmark problem within the OpenNeoMC framework designed specifically for reinforcement learning. The benchmark involves optimizing a unit cell of a research reactor with two varying parameters (fuel density and water spacing) to maximize neutron flux while maintaining reactor criticality. The test case features distinct local optima, representing different physical regimes, thus posing a challenge for learning algorithms. Through extensive simulations utilizing evolutionary and neuroevolutionary algorithms, we demonstrate the effectiveness of reinforcement learning in navigating complex optimization landscapes with strict constraints. Furthermore, we propose acceleration techniques within the OpenNeoMC framework, including model updating and cross-section usage by RAM utilization, to expedite simulation times. Our findings emphasize the importance of machine learning integration in reactor optimization and contribute to advancing methodologies for addressing intricate optimization challenges in nuclear engineering. The sources of this work are available at our GitHub repository: RLOpenNeoMC. © 2024, CC BY.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

RSVP for VPSA: A Meta Design Study on Rapid Suggestive visualization Prototyping for visual Parameter Space Analysis

arXiv

引用

arXiv 2024年

作者： Klaffenboeck, Manfred Gleicher, Michael Sorger, Johannes Wimmer, Michael Möller, Torsten The Institute of Visual Computing and Human Centered Technology TU Wien Austria The Complexity Science Hub Vienna Austria The Department of Computer Sciences The University of Wisconsin - Madison United States The Faculty of Computer Science The University of Vienna The Research Network Data Science Uni Wien Austria

visual Parameter Space Analysis (VPSA) enables domain scientists to explore input-output relationships of computational models. Existing VPSA applications often feature multi-view visualizations designed by visualization experts for a specific scenario, making it hard for domain scientists to adapt them to their problems without professional help. We present RSVP, the Rapid Suggestive visualization Prototyping system encoding VPSA knowledge to enable domain scientists to prototype custom visualization dashboards tailored to their specific needs. The system implements a task-oriented, multi-view visualization recommendation strategy over a visualization design space optimized for VPSA to guide users in meeting their analytical demands. We derived the VPSA knowledge implemented in the system by conducting an extensive meta design study over the body of work on VPSA. We show how this process can be used to perform a data and task abstraction, extract a common visualization design space, and derive a task-oriented VisRec strategy. User studies indicate that the system is user-friendly and can uncover novel insights. © 2024, CC BY.

关键词： visualization

来源：评论

学校读者我要写书评

暂无评论

Taking GPU Programming Models to Task for Performance Portability

arXiv

引用

arXiv 2024年

作者： Davis, Joshua H. Sivaraman, Pranav Kitson, Joy Parasyris, Konstantinos Menon, Harshitha Minn, Isaac Georgakoudis, Giorgis Bhatele, Abhinav Department of Computer Science University of Maryland United States Center for Applied Scientific Computing Lawrence Livermore National Laboratory United States

Portability is critical to ensuring high productivity in developing and maintaining scientific software as the diversity in on-node hardware architectures increases. While several programming models provide portability for diverse GPU platforms, they don’t make any guarantees about performance portability. In this work, we explore several programming models – CUDA, HIP, Kokkos, RAJA, OpenMP, OpenACC, and SYCL, to study if the performance of these models is consistently good across NVIDIA and AMD GPUs. We use five proxy applications from different scientific domains, create implementations where missing, and use them to present a comprehensive comparative evaluation of the programming models. We provide a Spack scripting-based methodology to ensure reproducibility of experiments conducted in this work. Finally, we attempt to answer the question – to what extent does each programming model provide performance portability for heterogeneous systems in real-world usage? Copyright © 2024, The Authors. All rights reserved.

关键词： Graphics processing unit

来源：评论

学校读者我要写书评

暂无评论

Accurately Estimating Unreported Infections using Information Theory

arXiv

引用

arXiv 2025年

作者： Cui, Jiaming Adhikari, Bijaya Haddadan, Arash Ul Haque, A.S.M. Ahsan Vreeken, Jilles Vullikanti, Anil Prakash, B. Aditya College of Computing Georgia Institute of Technology United States Department of Computer Science University of Virginia United States Department of Computer Science Virginia Tech United States Department of Computer Science The University of Iowa United States Amazon United States CISPA Helmholtz Center for Information Security Germany

One of the most significant challenges in combating against the spread of infectious diseases was the difficulty in estimating the true magnitude of infections. Unreported infections could drive up disease spread, making it very hard to accurately estimate the infectivity of the pathogen, therewith hampering our ability to react effectively. Despite the use of surveillance-based methods such as serological studies, identifying the true magnitude is still challenging. This paper proposes an information theoretic approach for accurately estimating the number of total infections. Our approach is built on top of Ordinary Differential Equations (ODE) based models, which are commonly used in epidemiology and for estimating such infections. We show how we can help such models to better compute the number of total infections and identify the parametrization by which we need the fewest bits to describe the observed dynamics of reported infections. Our experiments on COVID-19 spread show that our approach leads to not only substantially better estimates of the number of total infections but also better forecasts of infections than standard model calibration based methods. We additionally show how our learned parametrization helps in modeling more accurate what-if scenarios with non-pharmaceutical interventions. Our approach provides a general method for improving epidemic modeling which is applicable broadly. © 2025, CC BY-NC-ND.

关键词：

来源：评论

学校读者我要写书评

暂无评论

An IoT-Based Microservice Platform for Virtual Puppetry Performance

引用

IEEE Access 2023年 11卷 103014-103032页

作者： Luo, Helin Lin, Yi-Bing Liao, Chen-Chi Huang, Yung-Hui Academia Sinica Research Center for Information Technology Innovation Taipei115 Taiwan Tainan National University of the Arts Graduate Institute of Animation and Film Art Tainan720 Taiwan National Yang Ming Chiao Tung University Department of Computer Science and Engineering Hsinchu300 Taiwan China Medical University College of Humanities and Sciences Taichung406 Taiwan National Cheng Kung University Miin Wu School of Computing Tainan701 Taiwan Asia University Department of Computer Science and Information Engineering Taichung413 Taiwan

This study proposes a web-based real-time remote-controlled virtual puppetry (avatar) performance platform called AvatarTalk to combine traditional Taiwanese puppetry, the Internet of Things (IoT), and 3D avatar models. AvatarTalk enables puppeteers to control avatar puppets on the web using either motion capture gloves or camera-based image recognition. This transformative performance approach not only facilitates remote performances and multi-screen presentations across various devices but also introduces a fresh gesture interpretation technique for dance artists. AvatarTalk supports a mechanism that can accommodate new control and puppet devices from other approaches through the IoT-based microservice concept. We develop a calibration procedure to enable the accurate capture of hand gestures to manipulate the movements of virtual puppets, empowering them to perform fundamental traditional puppetry poses such as nodding, bowing, and synchronized hand movements. We have conducted experiments to show the accuracy of AvatarTalk calibration. Our study indicates that AvatarTalk can almost detect the right gestures (98.75%-100% recall) and very seldom mistake the wrong gestures (90.8%-100% precision). Additionally, we also provide the mechanism to measure the delays of controlling the puppets. An analytic model is proposed to design the delay times of the messages between the control device of a puppeteer and the AvatarTalk server. In the current AvatarTalk implementation, if the message delay does not exceed 0.1 seconds, four puppeteers can synchronize their actions if the elapsed time between two actions of a puppeteer is longer than 0.3 second. © 2013 IEEE.

关键词： Internet of things

来源：评论

学校读者我要写书评

暂无评论

Beatnik: A Novel Global Communication Mini-Application

arXiv

引用

arXiv 2024年

作者： Stewart, Jason R. Bridges, Patrick G. Department of Computer Science University of New Mexico United States Center for Advanced Research Computing University of New Mexico United States

Beatnik is a novel open source mini-application that exercises the complex communication patterns often found in production codes but rarely found in benchmarks or mini-applications. It simulates 3D Raleigh-Taylor instabilities based on Pandya and Shkoller's Z-Model formulation using the Cabana performance portability framework. This paper presents both the high-level design and important implementation details about Beatnik along with four benchmark setups for evaluating different aspects of HPC communication system performance. Evaluation results demonstrate both Beatnik's scalability on modern acceleratorbased systems using weak and strong scaling tests up to 1024 GPUs, along with Beatnik's ability to expose communication challenges in modern systems and solver libraries. © 2024, CC BY.

关键词： Benchmarking

来源：评论

学校读者我要写书评

暂无评论

Python And Opencv For Sign Language Recognition

Python And Opencv For Sign Language Recognition

引用

2023 IEEE International Conference on Device Intelligence, computing and Communication Technologies, DICCT 2023

作者： Srinivasan, R. Kavita, R. Kavitha, M. Mallikarjuna, Basetty Bhatia, Sandeep Agarwal, Bhawna Ahlawat, Vinay Goel, Amit Veltech University Chennai Dept of Computer Science and Engineering India Galgotias University School of Computing Science and Engineering Uttar Pradesh Greater Noida India Birla Institute of Technology Mesra Jaipur Center India Kiet Ncr Delhi Department of Cse Uttar Pradesh India

ISBN: (纸本)9781665474924

Hearing-impaired people cannot communicate with normal people easily. Most people are not aware of sign language recognition. To support this, machine learning and CV can be used to create an impact on the impaired. This can be improved into automatic editors, in which the person can easily understand the sign language of the impaired people by just using hand sign recognition. In non-verbal communication, hand gesture always being an important mode of communication and it plays a vital role to bridge gap between deaf and dumb people. Several sign language recognition systems have been developed but the systems are not flexible and cost-effective. Physically challenged people can express their emotions and feelings through sign language. In this paper, we develop a sign detector that can detect signs and numbers and also other signs that are used in sign language with the help of OpenCV and Keras modules in python. By using this technology, we can understand what they want to convey through sign language which is not a common language to communicate with people. OpenCV and Keras of python are the modules used to achieve our work and the proposed work proved to be a user-friendly approach to communication by using Python language to recognize sign languages for hearing-impaired persons. © 2023 IEEE.

关键词： Python

来源：评论

学校读者我要写书评

暂无评论

Causal discovery via conditional independence testing with proxy variables 24

Causal discovery via conditional independence testing with p...

引用

Proceedings of the 41st International Conference on Machine Learning

作者： Mingzhou Liu Xinwei Sun Yu Qiao Yizhou Wang School of Computer Science and Center on Frontiers of Computing Studies (CFCS) Peking University Sch. of Data Science Fudan University Dep. of Automation Shanghai Jiao Tong University Center on Frontiers of Computing Studies (CFCS) and School of Computer Science and Inst. for Artificial Intelligence and Nat'l Eng. Research Center of Visual Technology and Nat'l Key Lab of General Artificial Intelligence Peking University

Distinguishing causal connections from correlations is important in many scenarios. However, the presence of unobserved variables, such as the latent confounder, can introduce bias in conditional independence testing commonly employed in constraint-based causal discovery for identifying causal relations. To address this issue, existing methods introduced proxy variables to adjust for the bias caused by unobserveness. However, these methods were either limited to categorical variables or relied on strong parametric assumptions for identification. In this paper, we propose a novel hypothesis-testing procedure that can effectively examine the existence of the causal relationship over continuous variables, without any parametric constraint. Our procedure is based on discretization, which under completeness conditions, is able to asymptotically establish a linear equation whose coefficient vector is identifiable under the causal null hypothesis. Based on this, we introduce our test statistic and demonstrate its asymptotic level and power. We validate the effectiveness of our procedure using both synthetic and real-world data. Code is publicly available at https://***/lmz123321/proxy_causal_discovery.

关键词：

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：