检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Struckmeier, Oliver Redko, Ievgen Mallasto, Anton Arndt, Karol Heinonen, Markus Kyrki, Ville Aalto University Finland Intelligent Robotics Group Finland Noah's Ark Lab Huawei Technologies Hong Kong Department of Computer Science Finland

Optimal transport (OT) is a powerful geometric tool used to compare and align probability measures following the least effort principle. Despite its widespread use in machine learning (ML), OT problem still bears its computational burden, while at the same time suffering from the curse of dimensionality for measures supported on general high-dimensional spaces. In this paper, we propose to tackle these challenges using representation learning. In particular, we seek to learn an embedding space such that the samples of the two input measures become alignable in it with a simple affine mapping that can be calculated efficiently in closed-form. We then show that such approach leads to results that are comparable to solving the original OT problem when applied to the transfer learning task on which many OT baselines where previously evaluated in both homogeneous and heterogeneous DA settings. The code for our contribution is available at https://***/Oleffa/LaOT. Copyright © 2023, The Authors. All rights reserved.

关键词： Mapping

来源：评论

学校读者我要写书评

暂无评论

Self-supervised linear motion deblurring

arXiv

引用

arXiv 2020年

作者： Liu, Peidong Janai, Joel Pollefeys, Marc Sattler, Torsten Geiger, Andreas Computer Vision and Geometry Group Department of Computer Science ETH Zürich Switzerland Autonomous Vision Group Max Planck Institute for Intelligent Systems Univeristy of Tübingen Tübingen Germany Microsoft Mixed Reality and Artificial Intelligence Lab Zürich Switzerland Computer Vision and Medical Image Analysis Group Chalmers University of Technology Sweden

Motion blurry images challenge many computer vision algorithms, e.g., feature detection, motion estimation, or object recognition. Deep convolutional neural networks are state-of-the-art for image deblurring. However, obtaining training data with corresponding sharp and blurry image pairs can be difficult. In this paper, we present a differentiable reblur model for self-supervised motion deblurring, which enables the network to learn from real-world blurry image sequences without relying on sharp images for supervision. Our key insight is that motion cues obtained from consecutive images yield sufficient information to inform the deblurring task. We therefore formulate deblurring as an inverse rendering problem, taking into account the physical image formation process: we first predict two deblurred images from which we estimate the corresponding optical flow. Using these predictions, we re-render the blurred images and minimize the difference with respect to the original blurry inputs. We use both synthetic and real dataset for experimental evaluations. Our experiments demonstrate that self-supervised single image deblurring is really feasible and leads to visually compelling results. Both the code and datasets are available at https://***/ethliup/SelfDeblur. Copyright © 2020, The Authors. All rights reserved.

关键词： Image enhancement

来源：评论

学校读者我要写书评

暂无评论

Coral reef mosaicking using Teardrop and Fast Image Labeling

Coral reef mosaicking using Teardrop and Fast Image Labeling

引用

OCEANS - Europe

作者： Francis James Corpuz Prospero Naval Eusebio Capili Jaylord Jauod Roel John Judilla Maricor Soriano Vision and Image Processing Group Instrumentation Physics Laboratory National Institute of Physics Computer Vision and Machine Intelligence Group Department of Computer Science College of Engineering University of Philippines Quezon Philippines Computer Vision and Machine Intelligence Group Department of Computer Science College of Engineering University of Philippines Quezon Philippines Mapua Robotics Center Special Projects Laboratory Institutional Laboratory Mgmt Office Mapua Institute of Technology Manila Philippines Vision and Image Processing Group Instrumentation Physics Laboratory National Institute of Physics University of Philippines Quezon Philippines

In this paper, we present an inexpensive system for diverless video capture and fast image stitching of image frames for rapid reef assessment of shallow coral reefs. Our system has two main components: 1) Teardrop, a boat-towable, winged hull apparatus designed to house a commercial digital camera, and 2) a mosaicking algorithm to stitch the coral reef video into mosaics for further appreciation and analysis. The captured reef video is then separated into image frames which are to be stitched sequentially using Fast Image Labeling. The overlap between image frames is estimated using Single-Step DFT, an efficient sub-pixel estimation algorithm. The estimated overlap is used to compute for the area to be added to the mosaic space. The overlapping section between succeeding image-pairs are stitched along a seam determined by a minimal-cost path using dynamic programming. The visibility of the seam boundaries is further minimized by utilizing blending on multi-resolution splines. Experimental results on automated reef mosaics creation from actual coral reef video taken using Teardrop shows the performance of the system described. The main contribution of this work is the demonstration of a rapid reef visualization system using a diverless system and commercially available, non-research-grade imaging equipment.

关键词： Image segmentation Sea measurements Cameras Labeling Discrete Fourier transforms Sea floor

来源：评论

学校读者我要写书评

暂无评论

Nonverbal signaling for non-humanoid robots during human-robot collaboration 16

Nonverbal signaling for non-humanoid robots during human-rob...

引用

ACM/IEEE International Conference on Human-Robot Interaction (HRI)

作者： Elizabeth Cha Maja Matarić Terrence Fong Department of Computer Science University of Southern California Los Angeles CA Intelligent Robotics Group NASA Ames Research Center Mountain View CA

ISBN: (纸本)9781467383707

Non-humanoid robots are becoming increasingly utilized for collaborative tasks across many domains, including industrial and service settings. Collaborative tasks between the human and robot rely on each collaborator's ability to effectively convey their mental state while accurately estimating and interpreting their partner's knowledge, intent, and actions. My research focuses on nonverbal communication signals that a non-humanoid robot can utilize during human-robot collaboration. We focus on motion, light and sound as they are commonly used communication channels across many domains and are available on most robot platforms. As a first step towards this goal, I present a completed study exploring how to use a simple multimodal light and sound signal to request help during a collaborative task. We then discuss future work to generate and utilize more complex signals to convey a variety of statuses to improve collaboration.

关键词： Robot kinematics Collaboration Service robots Robot sensing systems Autonomous automobiles Vehicles

来源：评论

学校读者我要写书评

暂无评论

Traffic Sign Detection With Event Cameras and DCNN

arXiv

引用

arXiv 2022年

作者： Wzorek, Piotr Kryjak, Tomasz Embedded Vision Systems Group Computer Vision Laboratory Department of Automatic Control and Robotics AGH University of Science and Technology Kraków Poland Department of Digital Systems Silesian University of Technology Gliwice Poland

In recent years, event cameras (DVS – Dynamic vision Sensors) have been used in vision systems as an alternative or supplement to traditional cameras. They are characterised by high dynamic range, high temporal resolution, low latency, and reliable performance in limited lighting conditions – parameters that are particularly important in the context of advanced driver assistance systems (ADAS) and self-driving cars. In this work, we test whether these rather novel sensors can be applied to the popular task of traffic sign detection. To this end, we analyse different representations of the event data: event frame, event frequency, and the exponentially decaying time surface, and apply video frame reconstruction using a deep neural network called FireNet. We use the deep convolutional neural network YOLOv4 as a detector. For particular representations, we obtain a detection accuracy in the range of 86.9-88.9% mAP@0.5. The use of a fusion of the considered representations allows us to obtain a detector with higher accuracy of 89.9% mAP@0.5. In comparison, the detector for the frames reconstructed with FireNet is characterised by an accuracy of 72.67% mAP@0.5. The results obtained illustrate the potential of event cameras in automotive applications, either as standalone sensors or in close cooperation with typical frame-based cameras. © 2022, CC BY.

关键词： Cameras

来源：评论

学校读者我要写书评

暂无评论

An accurate 3D feature tracking system with wide-baseline stereo smart cameras

An accurate 3D feature tracking system with wide-baseline st...

引用

ACM/IEEE International Conference on Distributed Smart Cameras, ICDSC

作者： Dominik Rueß Kristian Manthey Ralf Reulke German AeroSpace Center Optical Information Systems Institute of Robotics and Mechatronics Berlin Germany Department of Computer Science Research and Lecture Group Computer Vision Humboldt-Universität zu Berlin Berlin Germany

A typical video surveillance system consists of at least one camera, controlled by an operator. To decrease the human error rate and to generally lessen the burden of operators, many object tracking systems have been implemented, most of which work in 2D image space. If used centralized, this is a very expensive task. Furthermore, if several views are to be fused, large inaccuracies arise due to ground plane assumptions, for instance. Lastly, in outdoor setups, quite often there is a need for slower channels like Wireless LAN which cannot cope with the full resolution data stream. We provide a smart camera system which performs the intensive tasks like background estimation or feature extraction. A central unit only has to process the received data in feature space, increasing scalability. Additionally, the object tracking problem is converted to an accurate 3D feature tracking, avoiding difficulties such as proper object segmentation and adding increased trajectory accuracy. The feature regions are computed within the smart camera. A wide-baseline feature matching approach has been employed to allow more freedom in the placement of the single smart cameras.

关键词： Trajectory Three dimensional displays Cameras Smart cameras Estimation Image reconstruction Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Visual guided robotic endodontic therapeutic system

Visual guided robotic endodontic therapeutic system

引用

International Conference on Information and Emerging Technologies, ICIET

作者： Tauseef Gulrez Abdul Karim Shahid Usman Sana Nadeem Ghafoor Chaudhary Intelligent Machines & Robotics Research Group Department of Computer Science Comsats Institute of Information Technology Lahore Pakistan Intelligent Machines & Robotics Research Group COMSATS Institute of Information Technology Lahore Pakistan Medical and Dental College The University of Lahore Pakistan

In this study, a visual-guided robotic endodontic therapeutic system was assessed, a specially designed visual servo controller model is also proposed in this system. Visual-guided robotic system will be mounted on the teeth within patient's mouth, while robot controller, and a root-canal image processor have shared control over its motion. With an on-line monitoring and positioning control, the multipurpose robotic system will perform automatic treatment procedures, including probing, drilling, filing, cleaning and filling. This paper will discuss the preliminary results on image processing based electrical torque control design of the proposed robotic system. The paper will also discuss the innovations involved from the traditional way endodontists treat root canal to science and technology based automation.

关键词： Irrigation Torque Robot kinematics Surgery Pixel Instruments

来源：评论

学校读者我要写书评

暂无评论

Augmented reality approach for paper map visualization

Augmented reality approach for paper map visualization

引用

International Conference on Communication and Computational Intelligence (INCOCCI)

作者： C Adithya K Kowsik D Namrata V S Nageli Sangeeta Shrivastava S Rakshit Department of Information Science B. M. Sreenivasaiah College of Engineering Bangalore India Computer Vision Group Center for Artificial Intelligence and Robotics Bangalore India

Paper map visualization is good value proposition, even in the `everything is digital' world, because of the huge information density and ease of use at any remote or mobile location, especially in the military context. Completely doing away with the paper map brings in a discomfort factor for the military user. In this paper, we present a paper map data visualization solution that augments the information from the digital world onto a spread of a paper map. Augmented Reality techniques are used to enhance paper maps digitally resulting in a 3D visualization space with some interactions. The dynamic information of the terrain, placing virtual objects and interaction are features of the digital world that are superimposed onto a paper map. We have used ARToolkit marker tracking based approach and a see through Head Mounted Device for implementation. The digital elevation data in VRML format is split into smaller bits before rendering. The rendering is done after culling the relevant part of the terrain data and the same is popped up in stereo for enhanced visualization. The user can place virtual objects using markers. The distance between markers (user and reference) is calculated and is used for interactive features. The interaction options are load/select, slice/split and latch/freeze. We have achieved better performance by loading only the region of interest and not the complete terrain. The paper also illustrates the basic hardware set-up required for implementation.

关键词： Augmented reality Data visualization Cameras Three dimensional displays Rendering (computer graphics) Solid modeling Training

来源：评论

学校读者我要写书评

暂无评论

Spatial-Temporal correlatons for unsupervised action classification

Spatial-Temporal correlatons for unsupervised action classif...

引用

Workshop on Motion and Video Computing

作者： Silvio Savarese Andrey DelPozo Juan Carlos Niebles Li Fei-Fei Beckman Institute University of Illinois Urbana-Champaign USA Department of Computer Science University of Illinois Urbana-Champaign USA Robotics and Intelligent Systems Group Universidad del Norte Colombia Department of Computer Science Princeton University USA

Spatial-temporal local motion features have shown promising results in complex human action classification. Most of the previous works [6],[16],[21] treat these spatial- temporal features as a bag of video words, omitting any long range, global information in either the spatial or temporal domain. Other ways of learning temporal signature of motion tend to impose a fixed trajectory of the features or parts of human body returned by tracking algorithms. This leaves little flexibility for the algorithm to learn the optimal temporal pattern describing these motions. In this paper, we propose the usage of spatial-temporal correlograms to encode flexible long range temporal information into the spatial-temporal motion features. This results into a much richer description of human actions. We then apply an unsupervised generative model to learn different classes of human actions from these ST-correlograms. KTH dataset, one of the most challenging and popular human action dataset, is used for experimental evaluation. Our algorithm achieves the highest classification accuracy reported for this dataset under an unsupervised learning scheme.

关键词： Humans Cameras Robustness Video sequences computer science intelligent robots Legged locomotion Feature extraction intelligent systems Trajectory

来源：评论

学校读者我要写书评

暂无评论

Monocular pedestrian detection: Survey and experiments

Monocular pedestrian detection: Survey and experiments

引用

作者： Enzweiler, Markus Gavrila, Dariu M. Department of Mathematics and Computer Science Image and Pattern Analysis Group University of Heidelberg Speyerer St. 4 69115 Heidelberg Germany Environment Perception Department Assistance Systems and Chassis Daimler AG Group Research Wilhelm Runge St. 11 89081 Ulm Germany Intelligent Systems Lab Faculty of Science University of Amsterdam Kruislaan 403 1098 SJ Amsterdam Netherlands

Pedestrian detection is a rapidly evolving area in computer vision with key applications in intelligent vehicles, surveillance, and advanced robotics. The objective of this paper is to provide an overview of the current state of the art from both methodological and experimental perspectives. The first part of the paper consists of a survey. We cover the main components of a pedestrian detection system and the underlying models. The second (and larger) part of the paper contains a corresponding experimental study. We consider a diverse set of state-of-the-art systems: wavelet-based AdaBoost cascade [74], HOG/ linSVM [11], NN/LRF [75], and combined shape-texture detection [23]. Experiments are performed on an extensive data set captured onboard a vehicle driving through urban environment. The data set includes many thousands of training samples as well as a 27-minute test sequence involving more than 20,000 images with annotated pedestrian locations. We consider a generic evaluation setting and one specific to pedestrian detection onboard a vehicle. Results indicate a clear advantage of HOG/ linSVM at higher image resolutions and lower processing speeds, and a superiority of the wavelet-based AdaBoost cascade approach at lower image resolutions and (near) real-time processing speeds. The data set (8.5 GB) is made public for benchmarking purposes. © 2009 IEEE.

关键词： Benchmarking

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：