检索结果-内蒙古大学图书馆

International Conference on computer Vision (ICCV)

作者： Bohai Gu Heng Fan Libo Zhang Institute of Software Chinese Academy of Sciences Beijing China University of Chinese Academy of Sciences Beijing China Department of Computer Science and Engineering University of North Texas Denton TX USA

Current arbitrary style transfer models are limited to either image or video domains. In order to achieve satisfying image and video style transfers, two different models are inevitably required with separate training processes on image and video domains, respectively. In this paper, we show that this can be precluded by introducing UniST, a Unified Style Transfer framework for both images and videos. At the core of UniST is a domain interaction transformer (DIT ), which first explores context information within the specific domain and then interacts contextualized domain information for joint learning. In particular, DIT enables exploration of temporal information from videos for the image style transfer task and meanwhile allows rich appearance texture from images for video style transfer, thus leading to mutual benefits. Considering heavy computation of traditional multi-head self-attention, we present a simple yet effective axial multi-head self-attention (AMSA) for DIT , which improves computational efficiency while maintains style transfer performance. To verify the effectiveness of UniST, we conduct extensive experiments on both image and video style transfer tasks and show that UniST performs favorably against state-of-the-art approaches on both tasks. Code is available at https://***/NevSNev/UniST.

关键词：

来源：评论

学校读者我要写书评

暂无评论

ICEBAR: Feedback-Driven Iterative Repair of Alloy Specifications 22

ICEBAR: Feedback-Driven Iterative Repair of Alloy Specificat...

引用

Proceedings of the 37th IEEE/ACM International Conference on Automated software engineering

作者： Simón Gutiérrez Brida Germán Regis Guolong Zheng Hamid Bagheri Thanhvu Nguyen Nazareno Aguirre Marcelo Frias Department of Computer Science University of Rio Cuarto Argentina and CONICET Argentina Department of Computer Science University of Rio Cuarto Argentina University of Nebraska-Lincoln United States of America George Mason University United States of America Department of Software Engineering Buenos Aires Institute of Technology Argentina and CONICET Argentina

ISBN: (纸本)9781450394758

Automated program repair (APR) techniques have shown great success in automatically finding fixes for programs in programming languages such as C or Java. In this work, we focus on repairing formal specifications, in particular for the Alloy specification language. As opposed to most APR tools, our approach to repair Alloy specifications, named ICEBAR, does not use test-based oracles for patch assessment. Instead, ICEBAR relies on the use of property-based oracles, commonly found in Alloy specifications as predicates and assertions. These property-based oracles define stronger conditions for patch assessment, thus reducing the notorious overfitting issue caused by using test-based oracles, typically observed in APR contexts. Moreover, as assertions and predicates are inherent to Alloy, whereas test cases are not, our tool is potentially more appealing to Alloy users than test-based Alloy repair tools. At a high level, ICEBAR is an iterative, counterexample-based process, that generates and validates repair candidates. ICEBAR receives a faulty Alloy specification with a failing property-based oracle, and uses Alloy’s counterexamples to build tests and feed ARepair, a test-based Alloy repair tool, in order to produce a repair candidate. The candidate is then checked against the property oracle for overfitting: if the candidate passes, a repair has been found; if not, further counterexamples are generated to construct tests and enhance the test suite, and the process is iterated. ICEBAR includes different mechanisms, with different degrees of reliability, to generate counterexamples from failing predicates and assertions. Our evaluation shows that ICEBAR significantly improves over ARepair, in both reducing overfitting and improving the repair rate. Moreover, ICEBAR shows that iterative refinement allows us to significantly improve a state-of-the-art tool for automated repair of Alloy specifications without any modifications to the tool.

关键词： Formal Specification Alloy Automated Repair

来源：评论

学校读者我要写书评

暂无评论

Attribute Based Federated-Reinforcement Learning Approach for Drone Authorization 1

引用

3rd International Conference on Machine Intelligence and Smart Systems, MISS 2023

作者： Rao, K. Rajesh Pradhan, Tribikram Krishna Prakasha, K. Bosch Global Software Technologies Private Ltd. Robert Bosch Bangalore560030 India Department of Computer Science and Engineering Tezpur University Assam Tezpur784028 India Department of Information and Communication Technology Manipal Institute of Technology MAHE Manipal576104 India

ISBN: (数字)9783031317231

ISBN: (纸本)9783031317224

Drones have become a vital part of the technology-driven world we live in today. From using it for delivering our day-to-day items to high grade military purposes, drones have spread their demands to various sectors of the economy. However, a little exploration has been done when it comes to talking about their security. Security generally comprises two aspects: authentication and authorization. While authentication is more to do with confirming the users with who they say they are, authorization is when the user’s permissions to access a resource is examined and a decision to approve or deny the access is taken by the system. This paper discusses authorization in detail using reinforcement learning and federated learning. The model is trained on a synthetic dataset, and is evaluated based on progressive validation loss (PVL), F1-Score, and a new metric called Permit Score. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

An exploratory study on automatic identification of assumptions in the development of deep learning frameworks

arXiv

引用

arXiv 2024年

作者： Yang, Chen Liang, Peng Ma, Zinan School of Artificial Intelligence Shenzhen Polytechnic University Shenzhen518055 China State Key Laboratory for Novel Software Technology Nanjing University Nanjing China School of Computer Science Wuhan University Wuhan430072 China Department of Computer Science and Engineering Chalmers University of Technology GothenburgS-412 96 Sweden

Context: Stakeholders constantly make assumptions in the development of deep learning (DL) frameworks. These assumptions are related to various types of software artifacts (e.g., requirements, design decisions, and technical debt) and can turn out to be invalid, leading to system failures. Existing approaches and tools for assumption management usually depend on manual identification of assumptions. However, assumptions are scattered in various sources (e.g., code comments, commits, pull requests, and issues) of DL framework development, and manually identifying assumptions has high costs (e.g., time and resources). Objective: The objective of the study is to evaluate different classification models for the purpose of identification with respect to assumptions from the point of view of developers and users in the context of DL framework projects (i.e., issues, pull requests, and commits) on GitHub. Method: First, we constructed a new and largest dataset (i.e., the AssuEval dataset) of assumptions collected from the TensorFlow and Keras repositories on GitHub. Then we explored the performance of seven non-transformers based models (e.g., Support Vector Machine, Classification and Regression Trees), the ALBERT model, and three decoder-only models (i.e., ChatGPT, Claude, and Gemini) for identifying assumptions on the AssuEval dataset. Results: The study results show that ALBERT achieves the best performance (f1-score: 0.9584) for identifying assumptions on the AssuEval dataset, which is much better than the other models (the 2nd best f1-score is 0.8858, achieved by the Claude 3.5 Sonnet model). Though ChatGPT, Claude, and Gemini are popular models, we do not recommend using them to identify assumptions in DL framework development because of their low performance. Fine-tuning ChatGPT, Claude, Gemini, or other language models (e.g., Llama3, Falcon, and BLOOM) specifically for assumptions might improve their performance for assumption identification. Conclusions: This stu

关键词： Chatbots

来源：评论

学校读者我要写书评

暂无评论

How to combine models? Principles and mechanisms to aggregate fuzzy cognitive maps

How to combine models? Principles and mechanisms to aggregat...

引用

Simulation Winter Conference

作者： Ryan Schuerkamp Philippe J. Giabbanelli Umberto Grandi Sylvie Doutre Department of Computer Science & Software Engineering Miami University Oxford OH USA Institut de Recherche en Informatique de Toulouse (IRIT) Université Toulouse Capitole Toulouse France

Fuzzy Cognitive Maps (FCMs) are graph-based simulation models commonly used to model complex systems. They are often built by participants and aggregated to compare the viewpoints of homogenous groups (e.g., anglers and ecologists) and increase the reliability of the FCM. However, the default approach for aggregation may propagate the errors of an individual participant, producing an aggregate FCM whose structure and simulation outcomes do not align with the system of interest. Alternative aggregation methods exist; however, there are no criteria to assess the quality of aggregation methods. We define nine desirable criteria for FCM aggregation algorithms and demonstrate how three existing aggregation procedures from social choice theory can aggregate FCMs and fulfill desirable criteria, enabling the assessment and comparison of FCM aggregation procedures to support modelers in selecting an aggregation algorithm. Moreover, we classify existing aggregation algorithms to provide structure to the growing body of aggregation approaches.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Automation of e-Workshop Project Control for Knowledge-Intensive Areas 7th

Automation of e-Workshop Project Control for Knowledge-Inten...

引用

7th International KES Conference on Smart Education and e-Learning, KES SEEL 2020

作者： Boldyreva, Elena A. Lisitsyna, Lubov S. Department of Software Engineering and Computer Science ITMO University Saint Petersburg Russia

ISBN: (纸本)9789811555831

This article discusses the issues of systematic processing of electronic workshops content to prepare competitive graduates, of various majors, in knowledge-intensive areas. The research proposes an approach to an automated workshop design for a specific student major considering numerous factors which influence the changes in the regulatory framework, student learning outcomes of the previous workshop passage, employers’ assessment, and the labor market needs. The analysis of expert assessment and the workshop content formation in this approach involves the implementation of the following stages, such as calculation of the confidence-in-experts degree, selection of subject learning tasks, development of a workshop project considering the relevance of each learning task, and the task complexity and evaluation (points) for each major. The result of this study is an information system for managing the project of an electronic workshop. It implements the approach to the formation of the actual practical part content of subjects for each major. The approach is based on the ADDIE model of pedagogical design. Using the information system, we have developed a workshop project of the major "Design of embedded systems" for the subject "Embedded systems." The workshop was tested by the fourth-year bachelor students. The workshop results were entered into the system, processed, and used to make recommendations for changing the content of the workshop. © 2020, The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Information systems

来源：评论

学校读者我要写书评

暂无评论

Comparative Analysis of Random Forest and Multilayer Perceptron Algorithms for Predicting Gross Domestic Product

Comparative Analysis of Random Forest and Multilayer Percept...

引用

International Conference on Emerging Trends in Networks and computer Communications (ETNCC)

作者： Ambrose A. Azeta Muyideen D. Adewale Patrick O. Omote Funmi O. Onaolapo Idowo A. Bakare Steven Tjiraso Department of Software Engineering Namibia University of Science and Technology Namibia. Africa Centre of Excellence on Technology Enhanced Learning National Open University of Nigeria Nigeria. Department of Computer and Mathematical Sciences Crawford University Ogun State Nigeria.

ISBN: (数字)9798350353266

ISBN: (纸本)9798350353273

This study examines the efficacy of machine learning techniques, particularly Multilayer Perceptron (MLP), and Random Forest and in forecasting Nigeria's Gross Domestic Product (GDP). Traditional statistical methods frequently fail to capture intricate, nonlinear relationships, thus necessitating advanced techniques. The dataset employed in this research encompasses 22 instances from 2000 to 2021. It includes variables such as GDP, healthcare expenditure, population, and the index of economic freedom obtained from the World Bank as well as the Nigerian Perceptions Index for corruption. The data was prepared to undergo training and testing following the CRISP-DM methodology. The models were trained and evaluated using a five-member cross-validation method. Performance metrics indicated that MLP outperformed Random Forest, with MAE values of 16.8228 and 25.3074 and RMSE values of 23.2111 and 33.2027, respectively. These findings suggest that MLP offers more accurate GDP predictions for Nigeria, effectively managing complex data relationships. This study provides valuable insights for enhancing predictive accuracy and supporting more effective economic planning and decision-making.

关键词： Training Machine learning algorithms Accuracy Economic indicators Biological system modeling Multilayer perceptrons Predictive models Prediction algorithms Random forests Testing

来源：评论

学校读者我要写书评

暂无评论

An Analysis of Depression Detection Techniques from Online Social Networks 2nd

An Analysis of Depression Detection Techniques from Online S...

引用

2nd International Conference on Intelligent Technologies and Applications, INTAP 2019

作者： Bilal, Uffaq Khan, Farhan Hassan Department of Computer Science and Software Engineering IIUI Islamabad Pakistan Islamabad Pakistan

ISBN: (纸本)9789811552311

Mental illness is caused by depression which may have a deep negative impact on individuals or on the society as a whole. It is a growing and severe problem which has the tendency to increase with time due to the extensive use of social networking websites such as Facebook, Twitter, Instagram etc. These social networking websites allow users to share images, videos, expressions and emotions. Depression is a form of mental illness. Patients suffering from depression have mood disorders such as low mood, high mood, lack of interest in things, etc. Machine learning techniques on text data and emojis have been applied to automatically classify a user into depressed and non-depressed. State-of-the-art classifiers have been used by the researcher to detect depressed individuals. Benchmark datasets are composed of text and emojis used in the social networking websites where classification is based on four factors Emotional Process, Temporal Process, Linguistic Style and combination of these three factors. In the proposed method, the emotional process is combined with their respective emojis to develop an automatic system for the detection of depressed patients. The features from the emotional process and emojis will be extracted and state-of-the-art classifiers have been proposed to be trained and evaluated using multiple classifiers using different combinations of part-of-speech tags. © 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

Games for Scrum Team Collaboration in the Global software Development Environment: A Literature Review

Games for Scrum Team Collaboration in the Global Software De...

引用

International Conference on Informatics and Computing (ICIC)

作者： Anita Hidayati Iklima Ermis Ismail Ade Rahma Yuly Henry Edison Department of Computer and Informatics Engineering Politeknik Negeri Jakarta Depok Indonesia Faculty of Computer Science Universitas Dian Nuswantoro Semarang Indonesia Department of Software Engineering Blekinge Institute of Technology Sweden Swedia

ISBN: (纸本)9798350345728

software development companies commonly use Global software Development (GSD) in their industry. A competent Scrum team supports the success of the GSD project. This research aims to identify the game components in the form of types, elements, and technologies that contribute to building experience for the Scrum team in a GSD environment. We used a Systematic Literature Review (SLR) with the Kitchenham and snowball method. From the 12 papers, we concluded that none of the games specifically focused on the competence of the Scrum team at GSD. The SLR results also identify ten game elements and technologies as a reference for game development to meet the challenges in GSD. The conclusion is that there are still great opportunities to develop a game that can improve the competence of the GSD Scrum team. Game development should refer to elements and identified technologies.

关键词： Industries Systematics Bibliographies Games software Teamwork Object recognition

来源：评论

学校读者我要写书评

暂无评论

Two Quantum Algorithms for Communication between Spacelike Separated Locations

SSRN

引用

SSRN 2022年

作者： Datta, Amitava Department of Computer Science and Software Engineering The University of Western Australia 35 Stirling Highway PerthWA6009 Australia

We propose two quantum algorithms for communication between two observers Alice and Bob, situated at spacelike separated locations. Alice and Bob share one qubit each of a Bell state . While sending classical information, Alice collapses the state of Bob’s qubit in two different ways depending on whether she wants to send 0 or 1. Bob’s first algorithm detects the classical bit with probability of error, but Alice and Bob need to share k Bell states for communicating a single classical bit. Bob’s second algorithm is more complex, but Bob can detect the classical bit deterministically using four ancilla qubits. We also discuss the possible reason why our algorithm can bypass the ‘No Communication’ theorem and possible applications of our algorithms. © 2022, The Authors. All rights reserved.

关键词： Separation

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：