检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Lew, Jaihyun Choi, Jooyoung Shin, Chaehun Jung, Dahuin Yoon, Sungroh Interdisciplinary Program in AI Seoul National University Korea Republic of Department of Electrical and Computer Engineering Seoul National University Korea Republic of School of Computer Science and Engineering Soongsil University Korea Republic of AIIS ASRI and INMC Seoul National University Korea Republic of

Video Frame Interpolation (VFI) aims to synthesize intermediate frames between existing frames to enhance visual smoothness and quality. Beyond the conventional methods based on the reconstruction loss, recent works have employed generative models for improved perceptual quality. However, they require complex training and large computational costs for pixel space modeling. In this paper, we introduce disentangled Motion Modeling (MoMo), a diffusion-based approach for VFI that enhances visual quality by focusing on intermediate motion modeling. We propose a disentangled two-stage training process. In the initial stage, frame synthesis and flow models are trained to generate accurate frames and flows optimal for synthesis. In the subsequent stage, we introduce a motion diffusion model, which incorporates our novel U-Net architecture specifically designed for optical flow, to generate bi-directional flows between frames. By learning the simpler low-frequency representation of motions, MoMo achieves superior perceptual quality with reduced computational demands compared to the generative modeling methods on the pixel space. MoMo surpasses state-of-the-art methods in perceptual metrics across various benchmarks, demonstrating its efficacy and efficiency in VFI. © 2024, CC BY-NC-SA.

关键词： Pixels

来源：评论

学校读者我要写书评

暂无评论

Search is All You Need for Few-shot Anomaly Detection

arXiv

引用

arXiv 2025年

作者： Wang, Qishan Guo, Jia Gao, Shuyong Wang, Haofen Xiong, Li Hu, Junjie Guo, Hanqi Zhang, Wenqiang Academy for Engineering and Technology Fudan University Shanghai China School of Biomedical Engineering Tsinghua University Beijing China School of Computer Science Fudan university Shanghai China College of Design & Innovation Tongji University Shanghai China School of Physics and Mechanical & Electrical Engineering Hexi University Gansu China Engineering Research Center of AI&Robotics Ministry of Education Academy for Engineering&Technology Fudan university Shanghai China

Few-shot anomaly detection (FSAD) has emerged as a crucial yet challenging task in industrial inspection, where normal distribution modeling must be accomplished with only a few normal images. While existing approaches typically employ multi-modal foundation models combining language and vision modalities for prompt-guided anomaly detection, these methods often demand sophisticated prompt engineering and extensive manual tuning. In this paper, we demonstrate that a straightforward nearest-neighbor search framework can surpass state-of-the-art performance in both single-class and multi-class FSAD scenarios. Our proposed method, VisionAD, consists of four simple yet essential components: (1) scalable vision foundation models that extract universal and discriminative features;(2) dual augmentation strategies - support augmentation to enhance feature matching adaptability and query augmentation to address the oversights of single-view prediction;(3) multi-layer feature integration that captures both low-frequency global context and high-frequency local details with minimal computational overhead;and (4) a class-aware visual memory bank enabling efficient one-for-all multi-class detection. Extensive evaluations across MVTec-AD, VisA, and Real-IAD benchmarks demonstrate VisionAD’s exceptional performance. Using only 1 normal images as support, our method achieves remarkable image-level AUROC scores of 97.4%, 94.8%, and 70.8% respectively, outperforming current state-of-the-art approaches by significant margins (+1.6%, +3.2%, and +1.4%). The training-free nature and superior few-shot capabilities of VisionAD make it particularly appealing for real-world applications where samples are scarce or expensive to obtain. Code is available at ***/Qiqigeww/VisionAD. Copyright © 2025, The Authors. All rights reserved.

关键词： Normal distribution

来源：评论

学校读者我要写书评

暂无评论

Goal-Conditioned Reinforcement Learning Approach for Autonomous Parking in Complex Environments

Goal-Conditioned Reinforcement Learning Approach for Autonom...

引用

International Conference on Artificial Intelligence in Information and Communication (ICAIIC)

作者： Taeyoung Kim Taemin Kang Seungah Son Kuk Won Ko Dongsoo Har CCS Graduate School of Mobility Korea Advanced Institute of Science and Technology(KAIST) Daejeon Rep. of Korea The Robotics Program Korea Advanced Institute of Science and Technology(KAIST) Daejeon Rep. of Korea Department of Future Mobility Engineering Halla University Wonju Rep. of Korea

ISBN: (数字)9798331507022

ISBN: (纸本)9798331507039

In complex urban environments, autonomous vehicle parking is challenged by irregularly parked cars, static obstacles, and narrow spaces. Traditional parking methods are often limited to well-structured parking lots and fail to adapt to unstructured scenarios like alleyways with obstacles and randomly positioned vehicles. This paper addresses this limitation by applying goal-conditioned reinforcement learning to enable autonomous vehicles to park in congested environments based on specified target positions and orientations. A custom simulation environment is developed using Pygame to simulate parking scenarios across four progressive levels of difficulty, with obstacles to test adaptability. Three RL algorithms, SAC, HER, and an improved HER variant, are implemented to compare the performance in the simulation environment. Experimental results demonstrate that the proposed approach significantly improves parking success rates and trajectory efficiency in complex scenarios, contributing to robust, adaptable parking solutions for autonomous vehicles in real-world applications.

关键词： Training Space vehicles Three-dimensional displays Urban areas Reinforcement learning Trajectory Reliability Vehicle dynamics Autonomous vehicles Testing

来源：评论

学校读者我要写书评

暂无评论

Scientometric Analysis of Artificial Intelligence Research in Agriculture 4

Scientometric Analysis of Artificial Intelligence Research i...

引用

4th International Conference on Informatics, Multimedia, Cyber and Information System, ICIMCIS 2022

作者： Maulana, Fairuz Iqbal Pramono, Andi Hamim, Miftahul Prihatin, Sukeipah Yuli Arifuddin, Rahman School of Computer Science Bina Nusantara University Computer Science Department Jakarta11480 Indonesia School of Design Bina Nusantara University Interior Design Department Jakarta11480 Indonesia Binus Business School Bina Nusantara University Entrepreneurship Business Creation Jakarta11480 Indonesia Politeknik Kota Malang Informatics Study Program Malang Indonesia University of Merdeka Malang Department of Electrical Engineering Malang Indonesia

ISBN: (纸本)9781665473279

Artificial Intelligence Research in Agriculture has grown so common, and the potential they provide is so revolutionary that it is considered essential for competitive growth. Until we conducted this research, only 1064 academic documents on Trends of Artificial Intelligence Research in Agriculture that we found from 2012 to 2021 were obtained by searching the Scopus database. This study uses bibliometric analysis methods to map scientific publications worldwide. We use an article selection process, the year limitation, then the database is exported to RIS and CSV format files. Researchers in the United States had the most published articles and were indexed by Scopus among the most prolific authors ($\mathrm{N}=287$), followed by the United Kingdom at second ($\mathrm{N}=139$) and Italy with 66 academic publications. Data analysis reveals an increasing trend in the number of Trends of Artificial Intelligence Research in Agriculture publications worldwide, as measured by the Scopus index. © 2022 IEEE.

关键词： Agriculture

来源：评论

学校读者我要写书评

暂无评论

BSH-Det3D: Improving 3D Object Detection with BEV Shape Heatmap

BSH-Det3D: Improving 3D Object Detection with BEV Shape Heat...

引用

IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

作者： You Shen Yunzhou Zhang Yanmin Wu Zhenyu Wang Linghao Yang Sonya Coleman Dermot Kerr College of Information Science and Engineering Northeastern University Shenyang China School of Electronic and Computer Engineering Peking University Shenzhen China Department of Electronic and Computer Engineering Faculty of Robotics and Engineering North-eastern University Technical University in Munich Germany Shenyang China Department of Electronic and Computer Engineering Technical University Munich Germany School of Computing Engineering and Intelligent Systems Ulster University N. Ireland UK

The progress of LiDAR-based 3D object detection has significantly enhanced developments in autonomous driving and robotics. However, due to the limitations of LiDAR sensors, object shapes suffer from deterioration in occluded and distant areas, which creates a fundamental challenge to 3D perception. Existing methods estimate specific 3D shapes and achieve remarkable performance. However, these methods rely on extensive computation and memory, causing imbalances between accuracy and real-time performance. To tackle this challenge, we propose a novel LiDAR-based 3D object detection model named BSH-Det3D, which applies an effective way to enhance spatial features by estimating complete shapes from a bird's eye view (BEV). Specifically, we design the Pillar-based Shape Completion (PSC) module to predict the probability of occupancy whether a pillar contains object shapes. The PSC module generates a BEV shape heatmap for each scene. After integrating with heatmaps, BSH-Det3D can provide additional information in shape deterioration areas and generate high-quality 3D proposals. We also design an attention-based densification fusion module (ADF) to adaptively associate the sparse features with heatmaps and raw points. The ADF module integrates the advantages of points and shapes knowledge with negligible overheads. Extensive experiments on the KITTI benchmark achieve state-of-the-art (SOTA) performance in terms of accuracy and speed, demonstrating the efficiency and flexibility of BSH-Det3D. The source code is available on https://***/mystorm16/BSH-Det3D.

关键词：

来源：评论

学校读者我要写书评

暂无评论

SimulFlow: Simultaneously Extracting Feature and Identifying Target for Unsupervised Video Object Segmentation

arXiv

引用

arXiv 2023年

作者： Hong, Lingyi Zhang, Wei Gao, Shuyong Lu, Hong Zhang, WenQiang Shanghai Key Laboratory of Intelligent Information Processing School of Computer Science Fudan University Shanghai China Engineering Research Center of AI & Robotics Ministry of Education Academy for Engineering & Technology Fudan University Shanghai China

Unsupervised video object segmentation (UVOS) aims at detecting the primary objects in a given video sequence without any human interposing. Most existing methods rely on two-stream architectures that separately encode the appearance and motion information before fusing them to identify the target and generate object masks. However, this pipeline is computationally expensive and can lead to suboptimal performance due to the difficulty of fusing the two modalities properly. In this paper, we propose a novel UVOS model called SimulFlow that simultaneously performs feature extraction and target identification, enabling efficient and effective unsupervised video object segmentation. Concretely, we design a novel SimulFlow Attention mechanism to bridege the image and motion by utilizing the flexibility of attention operation, where coarse masks predicted from fused feature at each stage are used to constrain the attention operation within the mask area and exclude the impact of noise. Because of the bidirectional information flow between visual and optical flow features in SimulFlow Attention, no extra hand-designed fusing module is required and we only adopt a light decoder to obtain the final prediction. We evaluate our method on several benchmark datasets and achieve state-of-the-art results. Our proposed approach not only outperforms existing methods but also addresses the computational complexity and fusion difficulties caused by two-stream architectures. Our models achieve 87.4% J &F on DAVIS-16 with the highest speed (63.7 FPS on a 3090) and the lowest parameters (13.7 M). Our SimulFlow also obtains competitive results on video salient object detection datasets. Copyright © 2023, The Authors. All rights reserved.

关键词： Optical flows

来源：评论

学校读者我要写书评

暂无评论

Convolutional Neural Networks for Image-Based High-Throughput Plant Phenotyping:A Review

引用

Plant Phenomics 2020年第1期2卷 73-94页

作者： Yu Jiang Changying Li Horticulture Section School of Integrative Plant ScienceCornell AgriTechCornell UniversityUSA School of Electrical and Computer Engineering College of EngineeringThe University of GeorgiaUSA Phenomics and Plant Robotics Center The University of GeorgiaUSA

Plant phenotyping has been recognized as a bottleneck for improving the efficiency of breeding programs,understanding plantenvironment interactions,and managing agricultural *** the past five years,imaging approaches have shown great potential for high-throughput plant phenotyping,resulting in more attention paid to imaging-based plant *** this increased amount of image data,it has become urgent to develop robust analytical tools that can extract phenotypic traits accurately and *** goal of this review is to provide a comprehensive overview of the latest studies using deep convolutional neural networks(CNNs)in plant phenotyping *** specifically review the use of various CNN architecture for plant stress evaluation,plant development,and postharvest quality *** systematically organize the studies based on technical developments resulting from imaging classification,object detection,and image segmentation,thereby identifying state-of-the-art solutions for certain phenotyping ***,we provide several directions for future research in the use of CNN architecture for plant phenotyping purposes.

关键词： Plant image thereby

来源：评论

学校读者我要写书评

暂无评论

RoboFold: A platform for Protein-folding Inspired Robot Swarms *

RoboFold: A platform for Protein-folding Inspired Robot Swar...

引用

WRC Symposium on Advanced robotics and Automation (WRC SARA)

作者： Feng Zhang Yuping Gu Shuai Yuan Yongliang Yang School of Electrical and Control Engineering Shenyang Jianzhu University Shenyang Liaoning China State Key Laboratory of Robotics Shenyang Institute of Automation Chinese Academy of Sciences Shenyang China Institutes for Robotics and Intelligent Manufacturing Chinese Academy of Sciences Shenyang Liaoning China School of Computer Science and Engineering and School of Electrical and Control Engineering Shenyang Jianzhu University Shenyang Liaoning China

Natural swarms have inspired various controlling algorithms for swarm robotics, while few of them were programmed in a similar way as the natural swarms. Defining a reliable programming method is still a daunting challenge in swarm robotics. In nature, the information of 3D protein structure is programmed in the sequence of amino acids of the protein. To mimic the programming method of protein folding, we present a platform for investigating protein-folding inspired programming method for robot swarms. This platform includes three parts: a vibration machine, protein-like robot sequences, and recording devices. The preliminary results presented here demonstrated the effectiveness of this platform and the feasibility of protein-folding inspired programming method for robot swarms.

关键词：

来源：评论

学校读者我要写书评

暂无评论

The Contribution of Pancasila Education and Civic Education on Value Education for The Students During COVID 19 Pandemic in Digital Era

The Contribution of Pancasila Education and Civic Education ...

引用

2022 IEEE International Conference on Electrical, computer, and Energy Technologies, ICECET 2022

作者： Benawa, Arcadius Lusia, Erma Alwino, Alfensius Irawan, Iwan Witono, Petrus Hepi Faculty of Economics and Communication Bina Nusantara University Character Building Development Center Mass Communication Program Communication Department Jakarta Indonesia Faculty of Economics and Communication Bina Nusantara University Tourism Department Jakarta Indonesia Bina Nusantara University School of Computer Science Character Building Development Center Computer Science Department Jakarta Indonesia Bina Nusantara University School of Industrial Engineering Industrial Engineering Department Jakarta Indonesia

ISBN: (数字)9781665470872

ISBN: (纸本)9781665470872

When all nations including Indonesia are hit by COVID-19 pandemic in era digital, we are compelled to conduct research with the main purpose, namely how much the contribution or influence of Pancasila Education and Civic Education on Value Education for students during the Covid-19 pandemic. The method used in this research is a quantitative method. The data was obtained by asking the students who were respondents to fill out a questionnaire. 150 students who have attended Pancasila and Citizenship courses were chosen by random sampling. Then the data obtained from the responses of the respondents were processed by the SPSS 22 program, and analyzed by path analysis. From the data processing and path analysis, the results of the research are obtained which prove that Pancasila Education has a significant contribution or influence on Value Education. Civic Education has no significant contribution or influence on Value Education. Based on these findings from this research, it is necessary to increase more intensive and serious efforts in delivering Pancasila Education, especially Civic Education in order to increase the contribution of both education to value education for the students, so that the students as the younger generation do not only have an understanding about Pancasila Education and Civic, but also be able to live the values, especially SPIRIT (Striving for the excellence, Perseverance, Integrity, Respect, Innovation, and Team Work) which are the core values at Bina Nusantara University. © 2022 IEEE.

关键词： Students

来源：评论

学校读者我要写书评

暂无评论

An Online Model-Free Reinforcement Learning Approach for 6-DOF Robot Manipulators

An Online Model-Free Reinforcement Learning Approach for 6-D...

引用

International Workshop on Robot Sensing (ROSE)

作者： Zeyad Hosny Abdullah Nassar Ahmed AboElyazeed Mahmoud Mohamed Mohammed Abouheaf Wail Gueaieb School of Electrical Engineering and Computer Science University of Ottawa Ottawa ON K1N6N5 Canada Robotics Engineering Bowling Green State University Bowling Green 43402 OH USA

Controlling 6 Degrees-of-Freedom (DoF) robotic manipulators in an online, model-free manner poses significant challenges due to their complex coupling, non-linearities, and the need to account for unmodeled dynamics. This paper introduces a model-free adaptive approach for real-time control of a 6 DoF “EPSON” robotic manipulator, without requiring any prior knowledge of the manipulator’s dynamics. Initially, we lay out the framework for an optimal control solution. A performance index is introduced, leveraging error dynamics and correction control signals, offering the capability to incorporate high-order error dynamics without the need to explicitly derive error trajectories. The order of error dynamics is determined by the chosen number of error samples. We assume a kernel-based solution structure aligning with the performance index, resulting in a temporal difference equation. This equation can be optimized to formulate a model-free control strategy. Subsequently, a reinforcement learning approach is adopted to approximate the underlying strategy. Infeasible exact solutions are overcome by employing a value iteration mechanism to adapt the actor-critic structures within an adaptive critics framework. To validate the proposed approach, it is compared against a conventional proportional-integral controller. A Unified Robot Description Format file is generated to facilitate the import of the robotic manipulator into the MATLAB Simulink environment, enabling its control. Ultimately, the proposed method yields superior results in terms of the dynamic characteristics of the response, demonstrating its effectiveness over the conventional approach.

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：