检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Tiboni, Gabriele Klink, Pascal Peters, Jan Tommasi, Tatiana D’Eramo, Carlo Chalvatzaki, Georgia Department of Control and Computer Engineering Politecnico di Torino Italy Department of Computer Science Technische Universität Darmstadt Germany Center for Artificial Intelligence and Data Science University of Würzburg Germany Darmstadt Germany Centre for Cognitive Science TU Darmstadt Germany Germany Center for Mind Brain and Behavior Uni. Marburg and JLU Giessen Germany

Varying dynamics parameters in simulation is a popular Domain Randomization (DR) approach for overcoming the reality gap in Reinforcement Learning (RL). Nevertheless, DR heavily hinges on the choice of the sampling distribution of the dynamics parameters, since high variability is crucial to regularize the agent’s behavior but notoriously leads to overly conservative policies when randomizing excessively. In this paper, we propose a novel approach to address sim-to-real transfer, which automatically shapes dynamics distributions during training in simulation without requiring real-world data. We introduce DOmain RAndomization via Entropy MaximizatiON (DORAEMON), a constrained optimization problem that directly maximizes the entropy of the training distribution while retaining generalization capabilities. In achieving this, DORAEMON gradually increases the diversity of sampled dynamics parameters as long as the probability of success of the current policy is sufficiently high. We empirically validate the consistent benefits of DORAEMON in obtaining highly adaptive and generalizable policies, i.e. solving the task at hand across the widest range of dynamics parameters, as opposed to representative baselines from the DR literature. Notably, we also demonstrate the Sim2Real applicability of DORAEMON through its successful zero-shot transfer in a robotic manipulation setup under unknown real-world parameters. Copyright © 2023, The Authors. All rights reserved.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Finding Place in a Design Space: Challenges for Supporting Community Design Efforts at Scale

引用

Proceedings of the ACM on Human-computer Interaction 2021年第CSCW1期5卷 1-30页

作者： MacNeil, Stephen Ding, Zijian Boone, Ashley Grubbs, Anthony Bryce Dow, Steven P. Dept. Of Cognitive Science University Of California San DiegoCA92122 United States Dept. Of Electrical And Computer Engineering University Of California San DiegoCA92122 United States Dept. Of Human-centered Design Engineering University Of Washington SeattleWA98195 United States

Many organizations have adopted design processes that integrate community voices to discover the real problems that communities face. Online discussion forums offer a familiar and flexible technology that can help facilitate discussion around problems and potential solutions. However, we lack understanding about what information community members share, how that information is structured, and how social interactions affect design processes at scale. This paper presents a mixed-methods analysis of Canvas, a learning management system, which enables users to contribute to the design of the platform by sharing and deliberating on problems and solutions in a discussion forum. We collected and analyzed 1412 ideas and 18,335 associated comments shared on the Canvas discussion forum. We found that the distributed nature of design information, the presence of duplicate ideas, and contributors' gaming behaviors made it difficult for the community to make sense of the design discussion. These gaming behaviors also constitute a new concern for participatory design research. Finally, we reflect on how Canvas community members contribute information to a shared design space and how future systems could more effectively coordinate community design efforts. © 2021 ACM.

关键词： Design

来源：评论

学校读者我要写书评

暂无评论

Graph Coordinates and Conventional Neural Networks - An Alternative for Graph Neural Networks

Graph Coordinates and Conventional Neural Networks - An Alte...

引用

IEEE International Conference on Big Data

作者： Zheyi Qin Randy Paffenroth Anura P. Jayasumana Electrical and Computer Engineering Colorado State University Fort Collins USA Department of Data Science Program Worcester Polytechnic Institute Worcester MA USA

Graph-based data present unique challenges and opportunities for machine learning. Graph Neural Networks (GNNs), and especially those algorithms that capture graph topology through message passing for neighborhood aggregation, have been a leading solution. However, these networks often require substantial computational resources and may not optimally leverage the information contained in the graph’s topology, particularly for large-scale or complex *** propose Topology Coordinate Neural Network (TCNN) and Directional Virtual Coordinate Neural Network (DVCNN) as novel and efficient alternatives to message passing GNNs, that directly leverage the graph’s topology, sidestepping the computational challenges presented by competing algorithms. Our proposed methods can be viewed as a reprise of classic techniques for graph embedding for neural network feature engineering, but they are novel in that our embedding techniques leverage ideas in Graph Coordinates (GC) that are lacking in current *** results, benchmarked against the Open Graph Benchmark Leaderboard, demonstrate that TCNN and DVCNN achieve competitive or superior performance to message passing GNNs. For similar levels of accuracy and ROC-AUC, TCNN and DVCNN need far fewer trainable parameters than contenders of the OGBN Leaderboard. The proposed TCNN architecture requires fewer parameters than any neural network method currently listed in the OGBN Leaderboard for both OGBN-Proteins and OGBN-Products datasets. Conversely, our methods achieve higher performance for a similar number of trainable parameters. These results hold across diverse datasets and edge features, underscoring the robustness and generalizability of our methods. By providing an efficient and effective alternative to message passing GNNs, our work expands the toolbox of techniques for graph-based machine learning. A significantly lower number of tunable parameters for a given evaluation metric makes TCNN and DVCNN especiall

关键词：

来源：评论

学校读者我要写书评

暂无评论

On nonrepetitive colorings of paths and cycles

arXiv

引用

arXiv 2023年

作者： Botler, Fábio Lomenha, Wanderson de Souza, João Pedro Program System Engineering and Computer Science COPPE Brazil Universidade Federal do Rio de Janeiro Brazil Colégio Pedro II Brazil

We say that a sequence a1 · · · a2t of integers is repetitive if ai = ai+t for every i ∈ {1, ..., t}. A walk in a graph G is a sequence v1 · · · vr of vertices of G in which vivi+1 ∈ E(...

We say that a sequence a1 · · · a2t of integers is repetitive if ai = ai+t for every i ∈ {1, ..., t}. A walk in a graph G is a sequence v1 · · · vr of vertices of G in which vivi+1 ∈ E(G) for every i ∈ {1, ..., r − 1}. Given a k-coloring c: V (G) → {1, ..., k} of V (G), we say that c is walk-nonrepetitive (resp. stroll-nonrepetitive) if for every t ∈ N and every walk v1 · · · v2t the sequence c(v1) · · · c(v2t) is not repetitive unless vi = vi+t for every i ∈ {1, ..., t} (resp. unless vi = vi+t for some i ∈ {1, ..., t}). The walk (resp. stroll) chromatic number σ(G) (resp. ρ(G)) of G is the minimum k for which G has a walk-nonrepetitive (resp. stroll-nonrepetitive) k-coloring. Let Cn and Pn denote, respectively, the cycle and the path with n vertices. In this paper we present three results that answer questions posed by Barát and Wood in 2008: (i) σ(Cn) = 4 whenever n ≥ 4 and n ∈/{5, 7};(ii) ρ(Pn) = 3 if 3 ≤ n ≤ 21 and ρ(Pn) = 4 otherwise;and (iii) ρ(Cn) = 4, whenever n ∈/{3, 4, 6, 8}, and ρ(Cn) = 3 otherwise. In particular, (ii) improves bounds on n obtained by Tao in *** Codes 05C15, 05C38 Copyright © 2023, The Authors. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Direct Machine Translation Indonesian-Batak Toba

Direct Machine Translation Indonesian-Batak Toba

引用

Conference on New Media Studies (CoNMedia)

作者： Zaenal Abidin Akmal Junaidi Wamiliana Frengky Marulitua Togatorop Imam Ahmad Ajeng Savitri Puspaningrum Doctoral Program of Mathematics and Natural Sciences Universitas Lampung Bandarlampung Indonesia Department of Computer Science Universitas Lampung Bandarlampung Indonesia Department of Mathematics Universitas Lampung Bandarlampung Indonesia Faculty of Engineering and Computer Science Universitas Teknokrat Indonesia Bandarlampung Indonesia

The translation application was developed using Python and Visual Studio Code as the primary programming environment. Technological progress has led to the younger generation forsaking the Batak language in their daily interactions, perceiving it as inconsequential to their social circle and contributing to the decline of regional languages. There is concern that regional languages, namely the Batak language, may lose significance or face extinction if this trend persists. This translation program is being implemented to aid in preserving regional languages, specifically the Batak language, enabling it to conform to contemporary circumstances. The development of this translation engine utilized the Batak language dictionary. The process of designing, coding, and testing is responsible for creating this direct machine translation. The BLEU score is employed to assess the caliber of machine translation. The study's findings indicate that the % BLEU value for ten sentences was 32.3%. This score falls within the category of sentences whose meaning can be understood, but the sentence structure still needs development. The direct machine translation developed in this Web platform is highly proficient and practical.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Efficient Diffusion-Driven Corruption Editor for Test-Time Adaptation

arXiv

引用

arXiv 2024年

作者： Oh, Yeongtak Lee, Jonghyun Choi, Jooyoung Jung, Dahuin Hwang, Uiwon Yoon, Sungroh Department of Electrical and Computer Engineering Seoul National University Korea Republic of Interdisciplinary Program in Artificial Intelligence Seoul National University Korea Republic of School of Computer Science and Engineering Soongsil University Korea Republic of Division of Digital Healthcare Yonsei University Korea Republic of

Test-time adaptation (TTA) addresses the unforeseen distribution shifts occurring during test time. In TTA, performance, memory consumption, and time consumption are crucial considerations. A recent diffusion-based TTA approach for restoring corrupted images involves image-level updates. However, using pixel space diffusion significantly increases resource requirements compared to conventional model updating TTA approaches, revealing limitations as a TTA method. To address this, we propose a novel TTA method that leverages an image editing model based on a latent diffusion model (LDM) and fine-tunes it using our newly introduced corruption modeling scheme. This scheme enhances the robustness of the diffusion model against distribution shifts by creating (clean, corrupted) image pairs and fine-tuning the model to edit corrupted images into clean ones. Moreover, we introduce a distilled variant to accelerate the model for corruption editing using only 4 network function evaluations (NFEs). We extensively validated our method across various architectures and datasets including image and video domains. Our model achieves the best performance with a 100 times faster runtime than that of a diffusion-based baseline. Furthermore, it is three times faster than the previous model updating TTA method that utilizes data augmentation, making an image-level updating approach more feasible. ∗ © 2024, CC BY.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Performance Improvement of Joint Space Distance Measurement for PIP joints

Performance Improvement of Joint Space Distance Measurement ...

引用

International Conference on Imaging, Signal Processing and Communication (ICISPC)

作者： Moeko Suzuki Tomio Goto Dept. of Creative Engineering Program Nagoya Institute of Technology Nagoya Japan Dept. of Computer Science Nagoya Institute of Technology Nagoya Japan

In order to analyze progresses of rheumatoid arthritis, we are developing an application to measure the distance between finger joints from X-ray images. In this paper, we focus on second joints, which are known to be prone to arthritis. We have proposed a method of filtering images and detecting edges, which has improved accuracy. Experimental results show that practical application of this system will reduce the burden on physicians and contribute to early detection of rheumatoid arthritis.

关键词：

来源：评论

学校读者我要写书评

暂无评论

An Application-Aware Clustering Strategy in Vehicular Ad Hoc Network

An Application-Aware Clustering Strategy in Vehicular Ad Hoc...

引用

International Conference on Computing, Electronics & Communications engineering (iCCECE)

作者： Chi-Hung Wei Chin-Chih Chang Chien-Hao Chu Sean Hsiao Ph.D. Program in Engineering Science Chung Hua University Hsinchu City Taiwan Department of Computer Science and Information Engineering Chung Hua University Hsinchu City Taiwan Bachelor Program in Applied Foreign Language and Smart Commerce Chung Hua University Hsinchu City Taiwan

VANET (Vehicular Ad Hoc Network) was developed to address the unique challenges of communication in a mobile vehicular environment. This technology can enhance driving safety and traffic control. Moreover, similar principles can be applied to unmanned aerial vehicles and robotics. When transmitting data in moving vehicles, it is challenging as access points and base stations keep changing, requiring handovers between different access points and base stations. Clustering is a popular approach used to address this challenge. Common clustering strategies are based on the proximity, communication range, the density, the application requirements of the moving vehicles, or hybrid of them. In this paper, we propose a clustering strategy based on the application sublayer in the vehicle communication protocol. If two computing devices on the move run the same application, they can be grouped together. The contribution of this research is to develop a clustering mechanism based on the application with very low overhead. We also validate the proposed method through simulations. The simulation results indicate that the proposed method is feasible.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Application of 3D Printing and Design in Healthcare: A Focus on Undergraduate Research to Attract Students to Bioengineering and Related Disciplines

Application of 3D Printing and Design in Healthcare: A Focus...

引用

Frontiers in Education (FIE) Conference

作者： Cindy M. Zelada-Quiroz Brian L. Reyes-Santiago Eduardo I. Ortiz-Rivera Bio-Engineering Graduate Program University of Puerto Rico at Mayagüez Mayagüez PR Department of Computer Science and Engineering University of Puerto Rico at Mayagüez Mayagüez PR Department of Electrical and Computer Engineering University of Puerto Rico at Mayagüez Mayagüez PR

ISBN: (数字)9798350351507

ISBN: (纸本)9798350363067

This research-to-practice paper describes developing and analyzing state-of-the-art smart boots created by combining CAD technology and advanced 3D printing techniques to attract students in bio-engineering and related fields. The primary objective of this innovative immobilization boot is to expedite fracture recovery phases through an ergonomic design to ensure optimal patient comfort during its use. Technological solutions are crucial in aiding the rehabilitation process for fractures caused by falls, heavy lifting, or rotational trauma. However, cost and comfort-related issues persist, underscoring the need for alternative approaches. This research addresses these challenges and delves into the broader implications of fracture treatment, catalyzing future projects and investigations in bioengineering. Additionally, this study serves as an educational tool that sparks the interest of high school and engineering students, promoting multidisciplinary collaboration in innovation. By involving students in specialized courses covering 3D design, human bone anatomy, biology, and materials science, this initiative empowers them to deepen their knowledge and develop new technologies to address bone injury problems. Material analyses include evaluating the type of material depending on the fracture site, such as PLA for printing and cotton and silicone gel for the midsection between the splint and the body. This research aims to advance our understanding of the type of fracture, the methods associated with their treatment, and tissue repair processes during bone callus formation. To summarize, this multidisciplinary approach drives advancements in bio-engineering and related fields, aiming to enhance patient outcomes and inspire students to pursue further research in bio-engineering and related fields. As part of this endeavor, a list of university-level courses based on the experience of the University of Puerto Rico at Mayaguez (UPRM), such as biology, bio-materials, 3D

关键词： Technological innovation Materials science and technology Three-dimensional displays Tissue engineering Medical services Three-dimensional printing Bones Mechanical factors Manufacturing Sparks

来源：评论

学校读者我要写书评

暂无评论

Effect of Parasitic Leakage Currents Associated with the Regrown Aperture of AlGaN/GaN Current Aperture Vertical Electron Transistors (CAVETs)

Effect of Parasitic Leakage Currents Associated with the Reg...

引用

IEEE Electron Devices Technology and Manufacturing Conference (EDTM)

作者： Sara Pouladi Jae Hyun Ryou Department of Mechanical Engineering Texas Center for Superconductivity at UH Advanced Manufacturing Institute University of Houston USA Department of Electrical and Computer Engineering Materials Science and Engineering Program University of Houston USA

We present a numerical simulation on the effect of leakage paths in the regrown GaN layer in the aperture and above the current blocking layer (CBL) of current aperture vertical electron transistor (CAVET) devices. Here, a 2D TCAD modeling is employed to simulate a CAVET device structure considering two main origins of parasitic leakage current from CBL/regrown-GaN interface and gate/regrown-GaN bulk and their degree of detrimental effect on the characteristics of AlGaN/GaN CAVETs.

关键词： Performance evaluation Apertures Logic gates Numerical simulation Wide band gap semiconductors Numerical models Leakage currents

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：