检索结果-内蒙古大学图书馆

Unified regularity measures for sample-wise learning and generalization

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Zhang, Chi Ma, Xiaoning Liu, Yu Wang, Le Su, Yuanqi Liu, Yuehu Institute of Artificial Intelligence and Robotics Xi’an Jiaotong University Xi’an China College of Artificial Intelligence Xi’an Jiaotong University Xi’an China School of Computer Science and Technology Xi’an Jiaotong University Xi’an China

Fundamental machine learning theory shows that different samples contribute unequally both in learning and testing processes. Contemporary studies on DNN imply that such sample difference is rooted on the distribution of intrinsic pattern information, namely sample regularity. Motivated by the recent discovery on network memorization and generalization, we proposed a pair of sample regularity measures for both processes with a formulation-consistent representation. Specifically, cumulative binary training/generalizing loss (CBTL/CBGL), the cumulative number of correct classifications of the training/testing sample within training stage, is proposed to quantize the stability in memorization-generalization process;while forgetting/mal-generalizing events, i.e., the mis-classification of previously learned or generalized sample, are utilized to represent the uncertainty of sample regularity with respect to optimization dynamics. Experiments validated the effectiveness and robustness of the proposed approaches for mini-batch SGD optimization. Further applications on training/testing sample selection show the proposed measures sharing the unified computing procedure could benefit for both tasks. Copyright © 2021, The Authors. All rights reserved.

关键词： Machine learning

Unlimited neighborhood interaction for heterogeneous trajectory prediction

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Zheng, Fang Wang, Le Zhou, Sanping Tang, Wei Niu, Zhenxing Zheng, Nanning Hua, Gang School of Software Engineering Xi'An Jiaotong University China Institute of Artificial Intelligence and Robotics Xi'An Jiaotong University China University of Illinois at Chicago United States School of Computer Science and Technology Xidian University China Wormpex AI Research

Understanding complex social interactions among agents is a key challenge for trajectory prediction. Most existing methods consider the interactions between pairwise traffic agents or in a local area, while the nature of interactions is unlimited, involving an uncertain number of agents and non-local areas simultaneously. Besides, they treat heterogeneous traffic agents the same, namely those among agents of different categories, while neglecting people's diverse reaction patterns toward traffic agents in different categories. To address these problems, we propose a simple yet effective Unlimited Neighborhood Interaction Network (UNIN), which predicts trajectories of heterogeneous agents in multiple categories. Specifically, the proposed unlimited neighborhood interaction module generates the fused-features of all agents involved in an interaction simultaneously, which is adaptive to any number of agents and any range of interaction area. Meanwhile, a hierarchical graph attention module is proposed to obtain category-to-category interaction and agent-to-agent interaction. Finally, parameters of a Gaussian Mixture Model are estimated for generating the future trajectories. Extensive experimental results on benchmark datasets demonstrate a significant performance improvement of our method over the state-of-the-art methods. © 2021, CC BY-NC-ND.

关键词： Trajectories

ARUSH:Automated Runtime Universal Scanner for Hybrid Cloud computing

学校读者我要写书评

暂无评论

ARUSH:Automated Runtime Universal Scanner for Hybrid Cloud c...

2021 IEEE Globecom Workshops, GC Wkshps 2021

作者： Luthra, Jatin Choudhary, Arjun Sharma, Abhishek Jayakody, Dushantha Nalin K. The LNM Institute of Information Technology Jaipur India Sardar Patel University of Police Security and Criminal Justice Centre for Cyber Security Jodhpur India Sri Lanka Technological Campus Centre for Telecommunication Research School of Engineering Padukka Sri Lanka School of Computer Science and Robotics National Research Tomsk Polytechnic University Tomsk Russia Sciences and Technologies Department Universidade Aut onoma de Lisboa Lisboa1169-023 Portugal Autonoma TechLab Lisboa1169-023 Portugal

ISBN: (纸本)9781665423908

With the shift of leading companies to the cloud, so comes the risk of data leaks and breaches because of many possible misconfigurations in the setup of various cloud computing environments. Though many solutions have come up, each is platform-specific, and none of them targets multiple or hybrid cloud platforms all together. This work proposes one such solution where three major public cloud platforms are supported, and the main focus is on the storage services provided by them. Since there is no such open-source solution that supports multiple cloud platforms, the program in this work is supposed to provide such a solution. Each cloud platform provides native SDK libraries for issuing API calls to the platform. These API calls aim to query the currently employed configuration for the storage services at the very least if not change. The solution can combine all the libraries into a single driver program. The program can facilitate security management of storage accounts across multiple as well hybrid cloud services and opens possibilities for automation and routine security and audit checks. © 2021 IEEE.

关键词： Libraries

Sample hardness based gradient loss for long-tailed cervical cell detection

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Liu, Minmin Li, Xuechen Gao, Xiangbo Chen, Junliang Shen, Linlin Wu, Huisi Computer Vision Institute School of Computer Science and Software Engineering Shenzhen University China Shenzhen Institute of Artificial Intelligence of Robotics of Society Shenzhen China Guangdong Key Laboratory of Intelligent Information Processing Shenzhen University Shenzhen518060 China National Engineering Laboratory for Big Data System Computing Technology Shenzhen University Shenzhen518060 China University of California Irvine United States

Due to the difficulty of cancer samples collection and annotation, cervical cancer datasets usually exhibit a long-tailed data distribution. When training a detector to detect the cancer cells in a WSI (Whole Slice Image) image captured from the TCT (Thinprep Cytology Test) specimen, head categories (e.g. normal cells and inflammatory cells) typically have a much larger number of samples than tail categories (e.g. cancer cells). Most existing state-of-the-art long-tailed learning methods in object detection focus on category distribution statistics to solve the problem in the long-tailed scenario, without considering the "hardness" of each sample. To address this problem, in this work we propose a Grad-Libra Loss that leverages the gradients to dynamically calibrate the degree of hardness of each sample for different categories, and re-balance the gradients of positive and negative samples. Our loss can thus help the detector to put more emphasis on those hard samples in both head and tail categories. Extensive experiments on a long-tailed TCT WSI image dataset show that the mainstream detectors, e.g. RepPoints, FCOS, ATSS, YOLOF, etc. trained using our proposed Gradient-Libra Loss, achieved much higher (7.8%) mAP than that trained using cross-entropy classification loss. Copyright © 2022, The Authors. All rights reserved.

关键词： Cytology

Distributed formation trajectory planning for multi-vehicle systems

学校读者我要写书评

暂无评论

Distributed formation trajectory planning for multi-vehicle ...

American Control Conference (ACC)

作者： Binh Nguyen Truong Nghiem Linh Nguyen Tung Nguyen Hung La Mehdi Sookhak Thang Nguyen Department of Engineering Texas A&M University–Corpus Christi Corpus Christi TX USA School of Informatics Computing and Cyber Systems Northern Arizona University Flagstaff AZ USA School of Engineering Information Technology and Physical Sciences Federation University Australia Churchill VIC Australia Department of Information Technology Uppsala University Uppsala Sweden Department of Computer Science and Engineering Advanced Robotics and Automation (ARA) Lab University of Nevada Reno NV USA Department of Computing Sciences Texas A&M University–Corpus Christi Corpus Christi TX USA

This paper addresses the problem of distributed formation trajectory planning for multi-vehicle systems with collision avoidance among vehicles. Unlike some previous distributed formation trajectory planning methods, our proposed approach offers great flexibility in handling computational tasks for each vehicle when the global formation of all the vehicles changes. It affords the system the ability to adapt to the computational capabilities of the vehicles. Furthermore, global formation constraints can be handled at any selected vehicles. Thus, any formation change can be effectively updated without recomputing all local formations at all the vehicles. To guarantee the above features, we first formulate a dynamic consensus-based optimization problem to achieve desired formations while guaranteeing collision avoidance among vehicles. Then, the optimization problem is effectively solved by ADMM-based or alternating projection-based algorithms, which are also presented. Theoretical analysis is provided not only to ensure the convergence of our method but also to show that the proposed algorithm can surely be implemented in a fully distributed manner. The effectiveness of the proposed method is illustrated by a numerical example of a 9-vehicle system.

关键词：

Taming time-varying information asymmetry in fresh status acquisition

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Wang, Zhiyuan Gao, Lin Huang, Jianwei Department of Computer Science and Engineering Chinese University of Hong Kong Hong Kong School of Electronics and Information Engineering Harbin Institute of Technology Shenzhen China School of Science and Engineering Chinese University of Hong Kong Shenzhen Hong Kong Shenzhen Institute of Artificial Intelligence and Robotics for Society

Many online platforms are providing valuable real-time contents (e.g., traffic) by continuously acquiring the status of different Points of Interest (PoIs). In status acquisition, it is challenging to determine how frequently a PoI should upload its status to a platform, since they are self-interested with private and possibly time-varying preferences. This paper considers a general multi-period status acquisition system, aiming to maximize the aggregate social welfare and ensure the platform freshness. The freshness is measured by a metric termed age of information. For this goal, we devise a long-term decomposition (LtD) mechanism to resolve the time-varying information asymmetry. The key idea is to construct a virtual social welfare that only depends on the current private information, and then decompose the per-period operation into multiple distributed bidding problems for the PoIs and platforms. The LtD mechanism enables the platforms to achieve a tunable trade-off between payoff maximization and freshness conditions. Moreover, the LtD mechanism retains the same social performance compared to the benchmark with symmetric information and asymptotically ensures the platform freshness conditions. Numerical results based on real-world data show that when the platforms pay more attention to payoff maximization, each PoI still obtains a non-negative payoff in the long-term. © 2021, CC BY.

关键词： Economic and social effects

COSMIC: fast closed-form identification from large-scale data for LTV systems

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Carvalho, Maria Soares, Claudia Lourenco, Pedro Ventura, Rodrigo GMV Portugal Instituto Superior Técnico U. de Lisboa Portugal NOVA LINCS Computer Science Department Nova School of Science and Technology U. Nova de Lisboa Portugal Institute for Systems and Robotics Lisboa Portugal

We introduce a closed-form method for identification of discrete-time linear time-variant systems from data, formulating the learning problem as a regularized least squares problem where the regularizer favors smooth solutions within a trajectory. We develop a closed-form algorithm with guarantees of optimality and with a complexity that increases linearly with the number of instants considered per trajectory. The COSMIC algorithm achieves the desired result even in the presence of large volumes of data. Our method solved the problem using two orders of magnitude less computational power than a general purpose convex solver and was about 3 times faster than a Stochastic Block Coordinate Descent especially designed method. Computational times of our method remained in the order of magnitude of the second even for 10k and 100k time instants, where the general purpose solver crashed. To prove its applicability to real world systems, we test with spring-mass-damper system and use the estimated model to find the optimal control path. Our algorithm was applied to both a Low Fidelity and Functional Engineering Simulators for the Comet Interceptor mission, that requires precise pointing of the on-board cameras in a fast dynamics environment. Thus, this paper provides a fast alternative to classical system identification techniques for linear time-variant systems, while proving to be a solid base for applications in the Space industry and a step forward to the incorporation of algorithms that leverage data in such a safety-critical environment. © 2021, CC BY.

关键词： Accident prevention

3d-Swim: 3d Vision Based Seam Width Measurement for Industrial Composite Fiber Layup In-Situ Inspection

学校读者我要写书评

暂无评论

SSRN

SSRN 2022年

作者： Zhang, Yinlong Yuan, Libiao Liang, Wei Xia, Xiaofang Pang, Zhibo State Key Laboratory of Robotics Shenyang Institute of Automation Chinese Academy of Sciences Shenyang110016 China Key Laboratory of Networked Control Systems Chinese Academy of Sciences Shenyang110016 China Institutes for Robotics and Intelligent Manufacturing Chinese Academy of Sciences Shenyang110169 China Shenyang Pharmaceutical University China School of Computer Science and Technology Xidian University Xian710071 China Department of Automation Technology ABB Corporate Research Sweden Vasteras72178 Sweden Stockholm11428 Sweden

Composite fiber layup in-situ inspection plays an important role in the advanced manufacturing. An accurate seam width measurement is an indispensable procedure during the layup inspection. However, the traditional seam measurements may fail in presence of surface curvature variance, curved edges between sheets, to name a few. This paper intentionally designs a novel 3D vision based Seam Width Measurement for Industrial Composite Fiber Layup In-situ inspection (called 3D-SWiM). It aims to address three major challenges in measuring the seam width between the composite fiber sheets, namely, seam region over-segmentation, failures on region of interests (ROI) seam detection, seam width measurement inconsistency. Firstly, the region growing with semantic refinement is designed for 3D point cloud clustering, to overcome the over-segmentation on seam regions. Afterwards, the gridding model with edge discrimination is applied to extract the 3D potential candidates in the edge areas. Eventually, we use the cubic B-spline fitting model to describe the seam region curves and the improved particle swarm optimization model to estimate the minimum and maximum distance between the seam edges. Our method has been evaluated extensively on the developed composite fiber layup platform. The experimental results prove the competitive accuracy against the state-of-the-art methods in seam width measurement. © 2022, The Authors. All rights reserved.

关键词： Inspection

Safer Motion Planning of Steerable Needles via a Shaft-to-Tissue Force Model

学校读者我要写书评

暂无评论

Journal of Medical robotics Research 2023年第1-2期8卷

作者： Bentley, Michael Rucker, Caleb Reddy, Chakravarthy Salzman, Oren Kuntz, Alan Robotics Center and Kahlert School of Computing University of Utah Salt Lake City 84112 UT United States The Department of Mechanical Aerospace and Biomedical Engineering University of Tennessee Knoxville 37996 TN United States Huntsman Cancer Institute and School of Medicine University of Utah Salt Lake City 84112 UT United States Department of Computer Science Technion-Israel Institute of Technology Technion City Haifa 3200003 Israel

Steerable needles are capable of accurately targeting difficult-to-reach clinical sites in the body. By bending around sensitive anatomical structures, steerable needles have the potential to reduce the invasiveness of many medical procedures. However, inserting these needles with curved trajectories increases the risk of tissue damage due to perpendicular forces exerted on the surrounding tissue by the needle's shaft, potentially resulting in lateral shearing through tissue. Such forces can cause significant tissue damage, negatively affecting patient outcomes. In this work, we derive a tissue and needle force model based on a Cosserat string formulation, which describes the normal forces and frictional forces along the shaft as a function of the planned needle path, friction model and parameters, and tip piercing force. We propose this new force model and associated cost function as a safer and more clinically relevant metric than those currently used in motion planning for steerable needles. We fit and validate our model through physical needle robot experiments in a gel phantom. We use this force model to define a bottleneck cost function for motion planning and evaluate it against the commonly used path-length cost function in hundreds of randomly generated three-dimensional (3D) environments. Plans generated with our force-based cost show a 62% reduction in the peak modeled tissue force with only a 0.07% increase in length on average compared to using the path-length cost in planning. Additionally, we demonstrate planning with our force-based cost function in a lung tumor biopsy scenario from a segmented computed tomography (CT) scan. By directly minimizing the modeled needle-to-tissue force, our method may reduce patient risk and improve medical outcomes from steerable needle interventions. © 2023 World Scientific Publishing Company.

关键词： bottleneck cost function continuum robots Force modeling motion planning robotic needle steering