检索结果-内蒙古大学图书馆

24th Working Notes of the Conference and Labs of the Evaluation Forum, CLEF-WN 2023

作者： Wolf, Stefan Beyerer, Jürgen Vision and Fusion Lab Karlsruhe Institute of Technology KIT c/o Technologiefabrik Haid-und-Neu-Str. 7 Karlsruhe76131 Germany Fraunhofer IOSB Institute of Optronics System Technologies and Image Exploitation Fraunhoferstrasse 1 Karlsruhe76131 Germany Fraunhofer Center for Machine Learning Germany

Fine-grained fungi species classification is an important task to support distinguishing edible and poisonous fungi and thus, reducing the risk of accidental poisoning. Therefore, the FungiCLEF 2023 challenge seeks to find the best solution for this task considering multiple metrics with each having a different application in focus like e.g., a low confusion of edible and poisonous fungi. We propose a method to approach the different metrics by exploiting modern deep learning networks, strong data augmentation and class-balanced training. The challenge assumes an open-set scenario which includes unknown classes during evaluation which we identify by a confidence thresholding approach. With our method, we achieved the 2nd place in the challenge with good scores across all metrics. Code is available at: https://***/wolfstefan/fungi2023. © 2023 Copyright for this paper by its authors.

关键词： Fungi

来源：评论

学校读者我要写书评

暂无评论

A PROOF OF CONCEPT: CYBERSECURITY LEARNING CONTENT IN LTI BASED, EDUCATIONAL systemS

A PROOF OF CONCEPT: CYBERSECURITY LEARNING CONTENT IN LTI BA...

引用

International Conference on Applied Computing 2022 and WWW/Internet 2022

作者： Kannegieser, Ehm Dyck, Sergius Atorf, Daniel Fraunhofer Institute of Optronics System Technologies and Image Interpretation IOSB Fraunhoferstr.1 Karlsruhe76131 Germany

ISBN: (纸本)9781713863793

Serious Games are an important part of technology aided learning. Since the learning outcome of intrinsically motivated learners is generally higher, a learning system should support the learner in achieving this state. Serious Games utilize the natural intrinsic motivation of playing a game by incorporating learning objectives. Concepts such as Digital Game Based Learning (DGBL) and Immersive Didactics achieve this by creating a learning environment that feels like an entertainment game by aligning learning and gaming goals. This puts players in a state of peaked intrinsic motivation which allows them to achieve increased levels of learning success - even while enjoying the game itself. Learning content for Serious Games is usually taken from existing courses or other materials. This work in progress explores the implementation of the interoperability standard LTI for integration purposes of third-party learning content into a Serious Game context. Copyright © (2022) by International Association for Development of the Information Society (IADIS). All rights reserved.

关键词： Interoperability

来源：评论

学校读者我要写书评

暂无评论

Concept of an autonomous in-situ sensor system for monitoring object movements caused by sea swells

Concept of an autonomous in-situ sensor system for monitorin...

引用

OCEANS

作者： Michael Wenrich Helge Renkewitz Sascha Kosleck Chair of Ocean Engineering University of Rostock Rostock Germany System Technologies and Image Exploitation IOSB-AST Fraunhofer Institute of Optronics Ilmenau Germany

ISBN: (数字)9798350362077

ISBN: (纸本)9798350362084

The concept of an autonomous in-situ sensor system, Motion Probe, for monitoring wave induced motions of underwater objects is presented. This cost-effective system integrates a pressure sensor alongside inertial and optionally global positioning sensors to collect data on the seastate, as well as the movement, acceleration, and orientation of the designated test object. Energy-efficient features, wireless charging, and Wi-Fi data transmission enhance user-friendly operation. An innovative housing design is introduced to protect the sensor from external influences, ensuring its reliability in marine environments. The paper outlines the electronic implementation, energy management, and pressure housing design, emphasizing the Motion Probe’s potential for various research applications. Testing phases include functional tests under controlled laboratory environments and long term sea trials, providing insights into wave inducted object dynamics.

关键词： Pressure sensors Oceans Laboratories Life estimation Reliability engineering Sensor systems Probes Monitoring Wireless fidelity Testing

来源：评论

学校读者我要写书评

暂无评论

Experimental separation estimation of incoherent optical sources reaching the (quantum) Cramér-Rao bound

Experimental separation estimation of incoherent optical sou...

引用

2023 Photonics North, PN 2023

作者： Rouviere, Clementine Barral, David Grateau, Antonin Sorelli, Giacomo Karuseichyk, Ilya Walschaers, Mattia Treps, Nicolas Sorbonne Université Collège de France Laboratoire Kastler Brossel ENS-Université Psl Cnrs 24 Rue Lhomond Paris75005 France Fraunhofer Iosb Ettlingen Fraunhofer Institute of Optronics System Technologies and Image Exploitation Gutleuthausstr. 1 Ettlingen76275 Germany

ISBN: (纸本)9798350326734

Resolving light sources below the diffraction limit is a fundamental task both for astronomy and microscopy. Several recent works analysed this problem through the lens of quantum parameter estimation theory and proved that the separation between two point sources can be estimated at the quantum limit using intensity measurements after spatial-mode demultiplexing. We implement this technique and provide an optimal estimator based on a linear combination of demultiplexed intensity measurements [1-3]. Our experimental setup allows for the generation of the images of two sources, with tunable mutual coherence, as well as for spatial mode demultiplexing to estimate their separation [4]. © 2023 IEEE.

关键词： Light sources

来源：评论

学校读者我要写书评

暂无评论

Adapting the Design of the Modal Holographic Wavefront Sensor to Present Atmospheric Turbulence Conditions

Adapting the Design of the Modal Holographic Wavefront Senso...

引用

Adaptive Optics and Applications, AOA 2022

作者： Zepp, Andreas Gladysz, Szymon Osten, Wolfgang Stein, Karin Fraunhofer IOSB Ettlingen Fraunhofer Institute of Optronics System Technologies and Image Exploitation Gutleuthausstraße 1 Ettlingen76275 Germany University of Stuttgart Institute for Applied Optics ITO Pfaffenwaldring 9 Stuttgart70569 Germany

ISBN: (纸本)9781557528209

The modal holographic wavefront sensor enables fast measurements of individual aberration modes without the need for time-consuming calculations. To minimize the effect of inter-modal crosstalk on the measurement accuracy, we present sensor optimization and show its performance in closed-loop adaptive optics in simulation. © 2022 The Author(s)

关键词： Atmospheric turbulence

来源：评论

学校读者我要写书评

暂无评论

Experimental separation estimation of incoherent optical sources reaching the Cramér-Rao bound

Experimental separation estimation of incoherent optical sou...

引用

2023 Conference on Lasers and Electro-Optics Europe and European Quantum Electronics Conference, CLEO/Europe-EQEC 2023

作者： Rouvière, Clémentine Barral, David Grateau, Antonin Sorelli, Giacomo Karuseichyk, Ilya Walschaers, Mattia Treps, Nicolas Laboratoire Kastler Brossel ENS-Université PSL CNRS Sorbonne Université Collège de France 24 Rue Lhomond Paris75005 France Fraunhofer IOSB Ettlingen Fraunhofer Institute of Optronics System Technologies and Image Exploitation Gutleuthausstr. 1 Ettlingen76275 Germany

ISBN: (纸本)9798350345995

Resolving light sources below the diffraction limit is a fundamental task both for astronomy and microscopy. Several recent works analysed this problem through the lens of quantum parameter estimation theory and proved that the separation between two point sources can be estimated at the quantum limit using intensity measurements after spatial-mode demultiplexing. We implement this technique and provide an optimal estimator based on a linear combination of demultiplexed intensity measurements [1-3]. Our experimental setup (see Fig. 1 (left)) allows for the generation of the images of two sources, with tunable mutual coherence, as well as for spatial mode demultiplexing to estimate their separation [4]. © 2023 IEEE.

关键词： Light sources

来源：评论

学校读者我要写书评

暂无评论

Towards a Stateless Cloud Native and Cloud Agnostic Digital Twin Platform for the Digital Product Passport: Out-of-the-box, Reliable, and Scaling

Towards a Stateless Cloud Native and Cloud Agnostic Digital ...

引用

International Conference on Emerging technologies and Factory Automation (ETFA)

作者： Magnus Redeker Sven Simikin Frank Marek Fraunhofer IOSB IOSB-INA Lemgo Fraunhofer Institute of Optronics System Technologies and Image Exploitation New York University Shanghai China Novatio-Solutions GmbH Germany

ISBN: (数字)9798350361230

ISBN: (纸本)9798350361247

Interoperable digital twins (DT) simplify and automate the implementation of Industry 4.0 (I4.0) use cases like Collaborative Condition Monitoring (CCM), Product and Corporate Carbon Footprint (PCF, CCF) and the Digital Product Passport (DPP) - aiming at adaptivity, resource-efficiency, user-friendliness, compliance with legal requirements and value-creation. Platforms accelerate the development and deployment of DTs and services and ensure their resilient operation in order to capture especially all relevant shop-floor information, for example to aggregate PCFs, CCFs, and DPPs or to report on sustainability - automatically. However, platform operators should in particular avoid cloud provider lock-ins to be able to react flexibly to use case requirements at any time. The technology must be manageable and costs must be controllable. The aim is to achieve a break-even point between security, costs, future viability and controllability, short implementation time and motivated use. Such requirements can currently only be met with cloud infrastructures. An agnostic design ensures that switching infrastructures is possible and can be done both securely and quickly. To this end, this paper develops an out-of-the-box cloud native and cloud agnostic DT platform providing specialized, stateless, loosely coupled DT services interacting within a service mesh. The developed platform meets the requirements, can easily be executed in any public, private and hybrid cloud, and can be scaled horizontally without any further ado.

关键词： Connectors Condition monitoring Costs Microservice architectures Switches Digital twins Servers Security Carbon footprint Portals

来源：评论

学校读者我要写书评

暂无评论

Attention-based Point Cloud Edge Sampling

arXiv

引用

arXiv 2023年

作者： Wu, Chengzhi Zheng, Junwei Pfrommer, Julius Beyerer, Jürgen Institute for Anthropomatics and Robotics Karlsruhe Institute of Technology Germany Fraunhofer Institute of Optronics System Technologies and Image Exploitation Iosb Germany

Point cloud sampling is a less explored research topic for this data representation. The most common sampling methods nowadays are still classical random sampling and farthest point sampling. With the development of neural networks, various methods have been proposed to sample point clouds in a task-based learning manner. However, these methods are mostly generative-based, other than selecting points directly with mathematical statistics. Inspired by the Canny edge detection algorithm for images and with the help of the attention mechanism, this paper proposes a non-generative Attention-based Point cloud Edge Sampling method (APES), which can capture the outline of input point clouds. Experimental results show that better performances are achieved with our sampling method due to the important outline information it learned. Copyright © 2023, The Authors. All rights reserved.

关键词： Sampling

来源：评论

学校读者我要写书评

暂无评论

A Cross Branch Fusion-Based Contrastive Learning Framework for Point Cloud Self-supervised Learning

A Cross Branch Fusion-Based Contrastive Learning Framework f...

引用

International Conference on 3D Imaging, Modeling, Processing, Visualization and Transmission (3DIMPVT)

作者： Chengzhi Wu Qianliang Huang Kun Jin Julius Pfrommer Jürgen Beyerer Institute for Anthropomatics and Robotics Karlsruhe Institute of Technology Germany Department of Automation Tsinghua University China System Technologies and Image Exploitation IOSB Fraunhofer Institute of Optronics Germany

ISBN: (数字)9798350362459

ISBN: (纸本)9798350362466

Contrastive learning is an essential method in self-supervised learning. It primarily employs a multi-branch strategy to compare latent representations obtained from different branches and train the encoder. In the case of multi-modal input, diverse modalities of the same object are fed into distinct branches. When using single-modal data, the same input undergoes various augmentations before being fed into different branches. However, all existing contrastive learning frameworks have so far only performed contrastive operations on the learned features at the final loss end, with no information exchange between different branches prior to this stage. In this paper, for point cloud unsupervised learning without the use of extra training data, we propose a Contrastive Cross-branch Attention-based framework for Point cloud data (termed PoCCA), to learn rich $3 D$ point cloud representations. By introducing sub-branches, PoCCA allows information exchange between different branches before the loss end. Experimental results demonstrate that in the case of using no extra training data, the representations learned with our self-supervised model achieve state-of-the-art performances when used for downstream tasks on point clouds.

关键词： Point cloud compression Representation learning Three-dimensional displays Training data Benchmark testing Data models Task analysis

来源：评论

学校读者我要写书评

暂无评论

Efficient Global Occupancy Mapping for Mobile Robots using OpenVDB

arXiv

引用

arXiv 2022年

作者： Hagmanns, Raphael Emter, Thomas Besselmann, Marvin Grosse Beyerer, Jürgen Fraunhofer Institute of Optronics System Technologies and Image Exploitation Karlsruhe Germany Fraunhofer Institute of Optronics System Technologies and Image Exploitation Karlsruhe Germany FZI Forschungszentrum Informatik Karlsruhe Germany

In this work we present a fast occupancy map building approach based on the VDB datastructure. Existing log-odds based occupancy mapping systems are often not able to keep up with the high point densities and framerates of modern sensors. Therefore, we suggest a highly optimized approach based on a modern datastructure coming from a computer graphic background. A multithreaded insertion scheme allows occupancy map building at unprecedented speed. Multiple optimizations allow for a customizable tradeoff between runtime and map quality. We first demonstrate the effectiveness of the approach quantitatively on a set of ablation studies and typical benchmark sets, before we practically demonstrate the system using a legged robot and a UAV. © 2022, CC BY.

关键词： Mapping

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：