检索结果-内蒙古大学图书馆

21st International Workshop of Physical Agents, WAF 2020

作者： Manso, Luis J. Jorvekar, Ronit R. Faria, Diego R. Bustos, Pablo Bachiller, Pilar Department of Computer Science College of Engineering and Physical Sciences Aston University Birmingham United Kingdom Department of Computer Engineering Pune Institute of Computer Technology Pune India Robotics and Artificial Vision Laboratory University of Extremadura Badajoz Spain

ISBN: (纸本)9783030625788

Autonomous navigation is a key skill for assistive and service robots. To be successful, robots have to comply with social rules, such as avoiding the personal spaces of the people surrounding them, or not getting in the way of human-to-human and human-to-object interactions. This paper suggests using Graph Neural Networks to model how inconvenient the presence of a robot would be in a particular scenario according to learned human conventions so that it can be used by path planning algorithms. To do so, we propose two automated scenario-to-graph transformations and benchmark them with different Graph Neural Networks using the SocNav1 dataset [1]. We achieve close-to-human performance in the dataset and argue that, in addition to its promising results, the main advantage of the approach is its scalability in terms of the number of social factors that can be considered and easily embedded in code in comparison with model-based approaches. The code used to train and test the resulting graph neural network is available in a public repository. © 2021, The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Human robot interaction

来源：评论

学校读者我要写书评

暂无评论

Activation Template Matching Loss for Explainable Face Recognition 17

Activation Template Matching Loss for Explainable Face Recog...

引用

17th IEEE International Conference on Automatic Face and Gesture Recognition, FG 2023

作者： Lin, Huawei Liu, Haozhe Li, Qiufu Shen, Linlin Computer Vision Institute School of Computer Science and Softwre Enginnering Shenzhen University Shenzhen China Shenzhen Institute of Artificial Intelligence and Robotics for Society Shenzhen China Guangdong Key Laboratory of Intelligent Information Processing Shenzhen University Shenzhen China

ISBN: (纸本)9798350345445

Can we construct an explainable face recognition network able to learn a facial part-based feature like eyes, nose, mouth and so forth, without any manual annotation or additionalsion datasets? In this paper, we propose a generic Explainable Channel Loss (ECLoss) to construct an explainable face recognition network. The explainable network trained with ECLoss can easily learn the facial part-based representation on the target convolutional layer, where an individual channel can detect a certain face part. Our experiments on dozens of datasets show that ECLoss achieves superior explainability metrics, and at the same time improves the performance of face verification without face alignment. In addition, our visualization results also illustrate the effectiveness of the proposed ECLoss. © 2023 IEEE.

关键词： Face recognition

来源：评论

学校读者我要写书评

暂无评论

Geometry Constrained Weakly Supervised Object Localization 16th

Geometry Constrained Weakly Supervised Object Localization

引用

16th European Conference on computer vision, ECCV 2020

作者： Lu, Weizeng Jia, Xi Xie, Weicheng Shen, Linlin Zhou, Yicong Duan, Jinming Computer Vision Institute School of Computer Science and Software Engineering Shenzhen University Shenzhen China Shenzhen Institute of Artificial Intelligence and Robotics for Society Shenzhen China School of Computer Science University of Birmingham Birmingham United Kingdom Department of Computer and Information Science University of Macau Zhuhai China

ISBN: (纸本)9783030585730

We propose a geometry constrained network, termed GC-Net, for weakly supervised object localization (WSOL). GC-Net consists of three modules: a detector, a generator and a classifier. The detector predicts the object location defined by a set of coefficients describing a geometric shape (i.e. ellipse or rectangle), which is geometrically constrained by the mask produced by the generator. The classifier takes the resulting masked images as input and performs two complementary classification tasks for the object and background. To make the mask more compact and more complete, we propose a novel multi-task loss function that takes into account area of the geometric shape, the categorical cross-entropy and the negative entropy. In contrast to previous approaches, GC-Net is trained end-to-end and predict object location without any post-processing (e.g. thresholding) that may require additional tuning. Extensive experiments on the CUB-200-2011 and ILSVRC2012 datasets show that GC-Net outperforms state-of-the-art methods by a large margin. Our source code is available at https://***/lwzeng/GC-Net. © 2020, Springer Nature Switzerland AG.

关键词： Geometry

来源：评论

学校读者我要写书评

暂无评论

Object Recognition and Pose Estimation using Laser scans For Advanced Underwater Manipulation

Object Recognition and Pose Estimation using Laser scans For...

引用

Symposium on Autonomous Underwater Vehicle Technology (AUV)

作者： Khadidja Himri Roger Pi Pere Ridao Nuno Gracias Albert Palomer Narcis Palomeras Computer Vision and Robotics Institute Universitat de Girona Girona Spain

ISBN: (纸本)9781728102542

3D object recognition is an active research area in computer vision and robotics. The integration of spatial information with semantic knowledge has become an important task for robots in order to successfully perform autonomous intervention missions. This paper presents an approach for the recognition and pose estimation of underwater objects, with the goal of enabling autonomous underwater intervention in man-made structures. The methods are developed to be used with raw data consisting of 3D colorless point clouds collected by a fast laser scanner. The proposed approach contains two main phases: Object recognition from range data, and feature-based semantic SLAM. The first goal consists of recognizing different objects present in the scene. For this purpose, a recognition and pose estimation pipeline was developed enclosing different steps such as segmentation, identification, and estimation of the position and orientation for each targeted object. The second goal aims at improving the AUV navigation in an underwater environment by using the result of the recognition and pose estimation pipeline to feed a feature based SLAM algorithm. As the AUV moves along the trajectory, the SLAM algorithm builds a map, recognizes targeted objects and integrates them into this map, and localizes its position with respect to it. Compared to previous experimental results performed in a water tank, this paper emphasizes the importance of estimating the pose of the objects (namely the orientation), as a way of promoting the accuracy of the robot localization.

关键词： Three-dimensional displays Object recognition Simultaneous localization and mapping Databases Lasers Pipelines

来源：评论

学校读者我要写书评

暂无评论

BALANCED OPTICAL FLOW REFINEMENT BY BIDIRECTIONAL CONSTRAINT

BALANCED OPTICAL FLOW REFINEMENT BY BIDIRECTIONAL CONSTRAINT

引用

IEEE International Conference on Image Processing

作者： Jongwon Choi Hyeongwoo Kim Tae-Hyun Oh In So Kweon Robotics and Computer Vision Laboratory Korea Advanced Institute of Science and Technology

ISBN: (纸本)9781479957521

We present an efficient optical flow refinement approach based on a bidirectional flow consistency. Our method is an add-on component that improves the existing optical flow estimation to be balanced between forward and backward flows. Most of the state-of-the-art optical flow methods only consider unidirectional motion vectors from a source image to a target image, which can make the estimated flow inconsistent with its backward estimation. The inconsistency can be reduced by considering the bidirectional motion when the optical flow is estimated, but it would be very hard for most of the typical optical flow methods and impossible for some of them. To solve this problem, we propose a sampling-based optimization method for efficiently refining the optical flows with a bidirectional constraint. By evaluating on Middlebury benchmark and public large displacement datasets, we validate the effectiveness of our method quantitatively and qualitatively and for accuracy.

关键词： Optical flow Bidirectional constraint Particle filter-based optimization

来源：评论

学校读者我要写书评

暂无评论

Online motion planning for underwater inspection

Online motion planning for underwater inspection

引用

Symposium on Autonomous Underwater Vehicle Technology (AUV)

作者： Marc Carreras Juan David Hernández Eduard Vidal Narcís Palomeras Pere Ridao Computer Vision and Robotics Institute Universitat de Girona Girona Spain

ISBN: (纸本)9781509024438

This paper proposes the use of path planning algorithms for AUVs in applications where the robot needs to adapt online its trajectory for inspection or safety purposes. These algorithms generate trajectories under motion constraints, which can be followed without deviations, to ensure the safety even when passing close to obstacles. View planning algorithms are also combined to decide the movements to be executed to discover the unexplored seabed or target and to cover it with a camera or sonar. Online mapping with profiling sonars and online planning with fast sampling-based algorithms, allows the execution of missions without any previous knowledge of the 3D shape of the environment. Real 2D results with a torpedo-shaped AUV with hovering capabilities in an artificial harbour structure and natural rocky canyon demonstrate the feasibility of the approach for avoiding or inspecting the underwater environment. Simulated 3D results also show the capability of the approach to be extended in more real and challenging environments.

关键词： Vehicles Planning Pipelines Robot sensing systems Sonar navigation Sonar

来源：评论

学校读者我要写书评

暂无评论

Rapid shallow coastal coral reef mapping using the teardrop system

Rapid shallow coastal coral reef mapping using the teardrop ...

引用

OCEANS 2012 MTS/IEEE Hampton Roads Conference: Harnessing the Power of the Ocean

作者： Corpuz, Francis James Naval Jr., Prospero Capili, Eusebio Jauod, Jaylord Judilla, Roel John Soriano, Maricor Vision and Image Processing Group National Institute of Physics University of the Philippines Diliman Quezon City Philippines Computer Vision and Machine Intelligence Group Department of Computer Science University of the Philippines Diliman Quezon City Philippines Mapua Robotics Center Institutional Laboratory Mgmt Office Mapua Institute of Technology Intramuros Manila Philippines

ISBN: (纸本)9781467308298

We present a rapid shallow coastal reef mapping system using a previously developed inexpensive diver less video capture system coupled with a fast-image mosaicking technique, using an affordable GPS logging device. The system replaces the traditional manta tow system of assessing and monitoring large areas of coastal coral reefs and at the same time provides historical visual record for future references. The system utilizes Teardrop, a hollow, teardrop-shaped that houses a commercial underwater point-and-shoot camera. The camera is set to video mode and captures video of the reef while the Teardrop is being towed by a wooden fishing boat. The resulting video is parsed into frames and mosaicked by using fast image labeling. The approximate location of the mosaicked images is recorded by using an inexpensive GPS logger. The coordinates are embedded on the mosaicked images and overlaid on mapping software such as Google Earth for visual record and spatial information integration. The same points are used to compute for the total distance covered by the capturing process. The whole process from coral reef video capture to mosaicking the images and finally embedding the mosaics in mapping software can be done in as short as 1 day, providing coastal managers and local government units with a tool to rapidly visualize and monitor shallow coral reefs. © 2012 IEEE.

关键词： Mapping

来源：评论

学校读者我要写书评

暂无评论

Pose-Driven Deep Models for Person Re-Identification

arXiv

引用

arXiv 2018年

作者： Eberle, Andreas Computer Vision for Human-Computer Interaction Research Group Institute for Anthropomatics and Robotics Karlsruhe Institute of Technology

Person re-identification (re-id) is the task of recognizing and matching persons at different locations recorded by cameras with non-overlapping views. One of the main challenges of re-id is the large variance in person poses and camera angles since neither of them can be influenced by the re-id system. In this work, an effective approach to integrate coarse camera view information as well as fine-grained pose information into a convolutional neural network (CNN) model for learning discriminative re-id embeddings is introduced. In most recent work pose information is either explicitly modeled within the re-id system or explicitly used for pre-processing, for example by pose-normalizing person images. In contrast, the proposed approach shows that a direct use of camera view as well as the detected body joint locations into a standard CNN can be used to significantly improve the robustness of learned re-id embeddings. On four challenging surveillance and video re-id datasets significant improvements over the current state of the art have been achieved. Furthermore, a novel reordering of the MARS dataset, called X-MARS is introduced to allow cross-validation of models trained for single-image re-id on tracklet data. Copyright © 2018, The Authors. All rights reserved.

关键词： Embeddings

来源：评论

学校读者我要写书评

暂无评论

REAL-TIME vision BASED AUV NAVIGATION SYSTEM USING A COMPLEMENTARY SENSOR SUITE

引用

IFAC Proceedings Volumes 2007年第17期40卷 373-378页

作者： Jonathan Horgan Daniel Toal Pere Ridao Rafael Garcia Mobile and Marine Robotics Research Centre Department of Electronic and Computer Engineering University of Limerick Ireland Computer Vision and Robotics Group Institute of Informatics and Applications University of Girona Spain

This paper proposes a real-time navigation system for an AUV that takes advantage of the complementary performance of a sensor suite including a DVL, a compass, a depth sensor and altimeter sensors with a feature based motion estimator using vision. To allow for real-time performance of the vision based motion estimator a simple but fast correlation algorithm is used for feature matching. The compass and the depth sensors are used to bound the drift of the heading and depth estimations respectively. The altimeter is required in order to translate the feature displacements measured from the images into the metric displacements of the robot. While the robot must rely on DVL navigation above a certain altitude where vision is useless, DVL measurements can be complemented with higher frequency accurate motion estimates from the vision system when navigating close to the seafloor.

关键词： marine robotics robot vision Kalman filtering robot navigation

来源：评论

学校读者我要写书评

暂无评论

KINECT UNBIASED

KINECT UNBIASED

引用

IEEE International Conference on Image Processing

作者： Manuel Martinez Rainer Stiefelhagen Karlsruhe Institute of Technology Institute for Anthropomatics and Robotics (IAR) Computer Vision for Human-Computer Interaction Lab

ISBN: (纸本)9781479957521

Since its release, Kinect has been the de facto standard for low-cost RGB-D sensors. An infrared laser ray shot through an holographic diffraction grating projects a fixed dot pattern which is captured using an infrared camera. The pseudo-random pattern ensures that a simple block matching algorithm suffices to provide reliable depth estimates, allowing a cost-effective implementation. In this paper, we analyze the software limitations of Kinect's method, which allows us to propose algorithms that provide better precision. First, we analyze the dot pattern: we measure its pincushion distortion and its effect on the dot density, which is smaller towards the edges of the image. Then, we analyze the behavior of Block Matching algorithms, we show how Kinect's Block Matching implementation is;in general;limited by the dot density of the pattern, and a significant spatial bias is introduced as a result. We propose an efficient approach to estimate the disparity of each dot, allowing us to produce a point cloud with better spatial resolution than Block Matching algorithms.

关键词： Kinect RGB-D Pattern Bias

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：