检索结果-内蒙古大学图书馆

Robust 3D Point Clouds Classification based on Declarative Defenders

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Li, Kaidong Zhang, Tianxiao Zhong, Chuncong Zhang, Ziming Wang, Guanghui Department of Electrical Engineering & Computer Science University of Kansas 1520 West 15th St. LawrenceKS66045 United States Department of Electrical & Computer Engineering Worcester Polytechnic Institute 100 Institute Rd. WorcesterMA01609 United States Department of Computer Science Toronto Metropolitan University 350 Victoria Street TorontoONM5B 2K3 Canada

3D point cloud classification requires distinct models from 2D image classification due to the divergent characteristics of the respective input data. While 3D point clouds are unstructured and sparse, 2D images are structured and dense. Bridging the domain gap between these two data types is a non-trivial challenge to enable model interchangeability. Recent research using Lattice Point Classifier (LPC) highlights the feasibility of cross-domain applicability. However, the lattice projection operation in LPC generates 2D images with disconnected projected pixels. In this paper, we explore three distinct algorithms for mapping 3D point clouds into 2D images. Through extensive experiments, we thoroughly examine and analyze their performance and defense mechanisms. Leveraging current large foundation models, we scrutinize the feature disparities between regular 2D images and projected 2D images. The proposed approaches demonstrate superior accuracy and robustness against adversarial attacks. The generative model-based mapping algorithms yield regular 2D images, further minimizing the domain gap from regular 2D classification tasks. The source code is available at https://***/KaidongLi/***. © 2024, CC BY.

关键词： Generative adversarial networks

Adaptive Knowledge Graphs Enhance Medical Question Answering: Bridging the Gap Between LLMs and Evolving Medical Knowledge

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Rezaei, Mohammad Reza Fard, Reza Saadati Parker, Jayson L. Krishnan, Rahul G. Lankarany, Milad Department of Biomedical Engineering University of Toronto Canada Department of Computer Science Worcester Polytechnic Institute United States Department of Biology University of Toronto Mississauga Canada Department of Computer Science University of Toronto Canada Vector Institute Canada

Large Language Models (LLMs) have greatly advanced medical Question Answering (QA) by leveraging vast clinical data and medical literature. However, the rapid evolution of medical knowledge and the labor-intensive process of manually updating domain-specific resources can undermine the reliability of these systems. We address this challenge with Adaptive Medical Graph-RAG (AMG-RAG), a comprehensive framework that automates the construction and continuous updating of Medical Knowledge Graphs (MKGs), integrates Chain-of-Thought (CoT) reasoning, and retrieves current external evidence (e.g., PubMed, WikiSearch). By dynamically linking new findings and complex medical concepts, AMG-RAG not only boosts accuracy but also enhances interpretability for medical queries. Evaluations on the MEDQA and MEDMCQA benchmarks demonstrate the effectiveness of AMG-RAG, achieving an F1 score of 74.1% on MEDQA and an accuracy of 66.34% on MEDMCQA—surpassing both comparable models and those 10 to 100 times larger. Importantly, these improvements are achieved without increasing computational overhead, underscoring the critical impact of automated knowledge graph generation and external evidence retrieval in delivering up-to-date, trustworthy medical insights. © 2025, CC BY.

关键词： Knowledge graph

FIRE-3DV: Framework-Independent Rendering Engine for 3D Graphics using Vulkan

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Allison, Christopher John Zhou, Haoying Munawar, Adnan Kazanzides, Peter Barragan, Juan Antonio Khoury College of Computer Sciences Northeastern University BostonMA02115 United States Department of Computer Science Johns Hopkins University BaltimoreMD21218 United States Department of Robotics Engineering Worcester Polytechnic Institute WorcesterMA01608 United States

Interactive dynamic simulators are an accelerator for developing novel robotic control algorithms and complex systems involving humans and robots. In user training and synthetic data generation applications, high-fidelity visualizations from the simulation are essential. Yet, robotic simulators often limit their rendering algorithms to preserve real-time interaction with the simulation. Advancements in Graphics Processing Units (GPU) enable improved visualization without compromising performance. However, these advancements cannot be fully leveraged in simulation frameworks that use legacy graphics application programming interfaces (API) to interface with the GPU. This paper presents a performance-focused and lightweight rendering engine supporting the modern Vulkan graphics API that can be easily integrated with other simulation frameworks to enhance visualizations. To illustrate the proposed method, our engine is used to modernize the legacy rendering pipeline of the Asynchronous Multi-Body Framework (AMBF), a dynamic simulation framework used extensively for interactive robotics simulation development. This new rendering engine implements graphical features such as physically based rendering (PBR), anti-aliasing, and ray-traced shadows, significantly improving the image fidelity of AMBF. Computational experiments show that the engine can render a simulated scene with over seven million triangles while maintaining GPU computation times within two milliseconds. Copyright © 2024, The Authors. All rights reserved.

关键词： Anti-aliasing

Bumblebee-Inspired C-V2X Dynamic Spectrum Access Testbed Using OpenAirInterface 91

学校读者我要写书评

暂无评论

Bumblebee-Inspired C-V2X Dynamic Spectrum Access Testbed Usi...

91st IEEE Vehicular Technology Conference, VTC Spring 2020

作者： Gill, Kuldeep S. Heath, Kevin N. Chuke, Sreeshti Haider, Aneela Gegear, Robert J. Ryder, Elizabeth F. Wyglinski, Alexander M. Worcester Polytechnic Institute Department of Electrical and Computer Engineering WorcesterMA United States Worcester Polytechnic Institute Department of Biology Biotechnology WorcesterMA United States

ISBN: (纸本)9781728152073

Cellular Vehicle-to-everything (C-V2X) was introduced by 3GPP in Release 14, which supports direct vehicle-to-vehicle (V2V) communication using sidelink channels. There are two modes in C-V2X for supporting V2V communication: Mode 3 and Mode 4. Both modes support V2V communications but differ on their spectrum allocation methodology. In out-of-coverage scenario defined under Mode 4, vehicles reserve spectrum resources using semi-persistent scheduling to enable direct communication. Currently, Intelligent Transportation System (ITS) spectrum at 5.9 GHz is emerging as an immediate resource to facilitate vehicular communication which has already been assigned for IEEE 802.11p. However, these dedicated channels will not be sufficient to support the connected vehicles in densely populated cities. Leveraging the existing underutilized wireless spectrum elsewhere, such as in digital television spectrum band is a likely solution. An intelligent scheme is required which can make adaptive decisions based on channel environment to efficiently use the secondary spectrum. Bumblebee behavioral model possess evolved decision-making mechanisms to adaptively solve similar problems while foraging in environments containing multiple floral resources (channels). Simulation results using bumblebee-based algorithm have shown that a memory system based on the averaging of stored channel energy information dramatically increased VDSA performance over a memoryless system in both urban and highway scenarios by 52% and 20%, respectively. In this paper, we propose a vehicular dynamic spectrum access (VDSA) solution based on bumblebee behavioral model for C-V2X and evaluate its performance using OpenAirInterface (OAI) LTE-Sidelink platform. © 2020 IEEE.

关键词： Vehicle to Everything

A Survey on 3D LiDAR Localization for Autonomous Vehicles 31

学校读者我要写书评

暂无评论

A Survey on 3D LiDAR Localization for Autonomous Vehicles

31st IEEE Intelligent Vehicles Symposium, IV 2020

作者： Elhousni, Mahdi Huang, Xinming Worcester Polytechnic Institute Electrical and Computer Engineering Department WorcesterMA01609 United States

LiDAR sensors are becoming one of the most essential sensors in achieving full autonomy for self driving cars. LiDARs are able to produce rich, dense and precise spatial data, which can tremendously help in localizing and tracking a moving vehicle. In this paper, we review the latest finding in 3D LiDAR localization for autonomous driving cars, and analyse the results obtained by each method, in an effort to guide the research community towards the path that seems to be the most promising. © 2020 IEEE.

关键词： Autonomous vehicles

Contact Testing of an Impedance-based Cancer Detection Probe

学校读者我要写书评

暂无评论

Contact Testing of an Impedance-based Cancer Detection Probe

Impedance Spectroscopy (IWIS), International Workshop on

作者： Erik Howard Skorina Shu Guo Maqsood Mughal Cagdas D. Onal Ahmet C. Sabuncu Robotics Engineering Worcester Polytechnic Institute Worcester USA Mechanical Engineering Worcester Polytechnic Institute Worcester USA Electrical & Computer Engineering Worcester Polytechnic Institute Worcester USA

Colorectal cancer (CRC) is a third leading cause of cancer death in United States. Colonoscopy is widely utilized both to detect CRC and to detect and remove adenomatous polyps, precursor lesion to CRC, thus preventing CRC from developing. The current technique for polyp identification requires a biopsy, a process that is difficult to perform in vivo. We seek to improve colonoscopy process by developing an impedance analyzer probe which can detect differences in tissue within existing colonoscopy systems. To this end, we developed a coaxial impedance probe capable of fitting through the 3mm instrument channel of a colonoscope. Using porcine intestinal tissue, we verified that this probe is capable of consistently measuring impedance under different contact conditions, such as variable forces or noisy tissue contact.

关键词： Spectroscopy In vivo Impedance measurement Colonoscopy Impedance Noise measurement Lesions

Security for emerging miniaturized wireless biomedical devices: Threat modeling with application to case studies

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Vakhter, Vladimir Soysal, Betul Schaumont, Patrick Guler, Ulkuhan Department of Electrical and Computer Engineering Worcester Polytechnic Institute Worcester United States

The landscape of miniaturized wireless biomedical devices (MWBDs) is rapidly expanding as proactive mobile healthcare proliferates. MWBDs are diverse and include various injectable, ingestible, implantable, and wearable devices. While the growth of MWBDs increases the flexibility of medical services, the adoption of these technologies brings privacy and security risks for their users. MWBDs can operate with sensitive, private information and affect patients through the use of stimulation and drug delivery. Therefore, these devices require trust and need to be secure. Embedding protective mechanisms into MWBDs is challenging because they are restricted in size, power budget, as well as processing and storage capabilities. Nevertheless, MWBDs need to be at least minimally securable in the face of evolving threats. The main intent of this work is to make the primary stakeholders of MWBDs aware of associated risks and to help the architects and the manufacturers of MWBDs protect their emerging designs in a repeatable and structured manner. Making MWBDs securable begins with performing threat modeling. This paper introduces a domain-specific qualitative-quantitative threat model dedicated to MWBDs. The proposed model is then applied to representative case studies from each category of MWBDs. © 2021, CC BY.

关键词： Wearable technology

A review for design,mechanism,fabrication,and application of magnetically responsive microstructured functional surface

学校读者我要写书评

暂无评论

International Journal of Extreme Manufacturing 2025年第1期7卷 96-128页

作者： Jian Wang Xingyi Song Chaochao Wang Yumei Zhou Ri Chen Yong Yang Bin Liu Yihao Zheng Hui Li Wei Zhou Lelun Jiang Guangdong Provincial Key Laboratory of Sensor Technology and Biomedical Instrument School of Biomedical EngineeringShenzhen Campus of Sun Yat-sen UniversityShenzhen 518107People’s Republic of China School of Mechatronic Engineering Guangdong Polytechnic Normal UniversityGuangzhou 510665People’s Republic of China School of Mechanical and Electrical Engineering Guangdong Polytechnic of Industry and CommerceGuangzhou 510550People’s Republic of China Mechanical&Materials Engineering Department Worcester Polytechnic InstituteWorcesterMA 01609United States of America Guangdong Provincial Key Laboratory of Micro/Nano Optomechatronics Engineering College of Mechatronics and Control EngineeringShenzhen UniversityShenzhen 518060People’s Republic of China Pen-Tung Sah Institute of Micro-Nano Science and Technology Xiamen UniversityXiamen 361005People’s Republic of China

Magnetically responsive microstructured functional surface(MRMFS),capable of dynamically and reversibly switching the surface topography under magnetic actuation,provides a wireless,noninvasive,and instantaneous way to accurately control the microscale engineered *** the last decade,many studies have been conducted to design and optimize MRMFSs for diverse applications,and significant progress has been *** review comprehensively presents recent advancements and the potential prospects in *** first classify MRMFSs into one-dimensional linear array MRMFSs,two-dimensional planar array MRMFSs,and dynamic self-assembly MRMFSs based on their ***,an overview of three deformation mechanisms,including magnetically actuated bending deformation,magnetically driven rotational deformation,and magnetically induced self-assembly deformation,are *** main fabrication strategies employed to create MRMFSs are summarized,including replica molding,magnetization-induced self-assembly,laser cutting,and ferrofluid-infused ***,the applications of MRMFS in droplet manipulation,solid transport,information encryption,light manipulation,triboelectric nanogenerators,and soft robotics are ***,the challenges that limit the practical applications of MRMFSs are discussed,and the future development of MRMFSs is proposed.

关键词： functional surfaces microstructures magnetic actuation manipulation

Suturing Tasks Automation Based on Skills Learned From Demonstrations: A Simulation Study

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Zhou, Haoying Jiang, Yiwei Gao, Shang Wang, Shiyue Kazanzides, Peter Fischer, Gregory S. Department of Robotics Engineering Worcester Polytechnic Institute WorcesterMA United States Department of Computer Science Johns Hopkins University BaltimoreMD United States Laboratory for Computational Sensing and Robotics Johns Hopkins University BaltimoreMD United States

In this work, we develop an open-source surgical simulation environment that includes a realistic model obtained by MRI-scanning a physical phantom, for the purpose of training and evaluating a Learning from Demonstration (LfD) algorithm for autonomous suturing. The LfD algorithm utilizes Dynamic Movement Primitives (DMP) and Locally Weighted Regression (LWR), but focuses on the needle trajectory, rather than the instruments, to obtain better generality with respect to needle grasps. We conduct a user study to collect multiple suturing demonstrations and perform a comprehensive analysis of the ability of the LfD algorithm to generalize from a demonstration at one location in one phantom to different locations in the same phantom and to a different phantom. Our results indicate good generalization, on the order of 91.5%, when learning from more experienced subjects, indicating the need to integrate skill assessment in the future. © 2024, CC BY.

关键词： Demonstrations