检索结果-内蒙古大学图书馆

Automatic Target Recognition XXXI 2021

作者： Freudenmann, Lucas Sommer, Lars Schumann, Arne Fraunhofer Iosb Institute of Optronics System Technologies and Image Exploitation Fraunhoferstrasse 1 Karlsruhe76131 Germany Fraunhofer Center for Machine Learning

ISBN: (数字)9781510642966

ISBN: (纸本)9781510642959

The increased availability of UAVs poses possibilities for novel applications but also safety issues regarding mass events or safety-sensitive infrastructure. Thus, the demand for automated UAV detection systems to allow for an early alert generation is increasing. Such systems often rely on electro-optical imagery and deep learning to detect UAVs. However, the absence of a large and diverse dataset for training may result in an error-prone learning process. In this work, we investigate how far these issues can be mitigated without relying on extra data, which is often costly to obtain and annotate. We thus evaluate and demonstrate the impact of different data augmentation strategies to enhance our available training data. We evaluate how the different methods increase the robustness of several state of the art deep learning based detectors. Particularly, we focus our evaluation on the aspects of false alarms caused by distractor objects or by complex background. © COPYRIGHT SPIE. Downloading of the abstract is permitted for personal use only.

关键词： Drones

来源：评论

学校读者我要写书评

暂无评论

Iterative 3D Modeling of Pulse Generation in End-Pumped Ho3+:YAG Laser Resonators Utilizing Active Q-switching

Iterative 3D Modeling of Pulse Generation in End-Pumped Ho3+...

引用

Conference on Lasers and Electro-Optics Europe (CLEO EUROPE)

作者： Marius Rupp Katharina Goth Marc Eichhorn Christelle Kieleck Fraunhofer IOSB (Institute of Optronics System Technologies and Image Exploitation) Ettlingen Germany Institute of Control Systems Karlsruhe Institute of Technology Karlsruhe Germany

ISBN: (纸本)9798350345995

High power solid-state lasers operating at around 2.1 μ m are used for a variety of applications, ranging from medical treatments and countermeasures to the pumping of mid-infrared optical parametric oscillators. Especially for the latter application, end-pumped crystal lasers in Q-switched operation have shown great potential because of the high pulse peak powers achievable with these systems [1]. Ho 3+ :YAG is very suitable as a pump source for nonlinear materials like zinc germanium phosphide (ZGP), due to lower parasitic absorption at wavelengths greater than 2 μm. A simulation model can greatly assist with the development, optimization, and understanding of such laser resonators. The model we present in this work focuses on the simulation of laser pulse generation by active Q-switching and builds upon a continuous-wave (cw) model we previously presented in [2].

关键词： High power lasers Laser crystals Laser operation Laser resonators Optical parametric oscillators Q switched lasers

来源：评论

学校读者我要写书评

暂无评论

Power Limits of Compact Ho3+:YAG Laser Resonators with Homogeneously Doped and Segmented Crystals

Power Limits of Compact Ho3+:YAG Laser Resonators with Homog...

引用

Conference on Lasers and Electro-Optics Europe (CLEO EUROPE)

作者： Katharina Goth Michael Griesbeck Marius Rupp Madeleine Eitner Marc Eichhorn Christelle Kieleck Fraunhofer IOSB (Institute of Optronics System Technologies and Image Exploitation) Ettlingen Germany Institute of Control Systems Karlsruhe Institute of Technology Karlsruhe Germany

ISBN: (纸本)9798350345995

Laser light from Ho 3+ : YAG crystals in the $2\ \mu\mathrm{m}$ range has a variety of applications like remote sensing, medical applications, or the pumping of optical parametric oscillators for the generation of mid-infrared radiation. When working with this quasi-three-level laser material, a high pump intensity is required which can be generated in an end-pumped design. However, such a design leads to a large longitudinal temperature gradient resulting from the exponential absorption of the pump light. One approach to decrease the temperature peak at the front of the crystal and therefore reduce temperature-induced stresses in the crystal is the use of diffusion-bonded crystals consisting of multiple segments with increasing doping concentrations along the crystal axis [1].

关键词： Laser light Laser materials Laser resonators Optical parametric oscillators Optical pumping Remote sensing

来源：评论

学校读者我要写书评

暂无评论

Quantum-inspired exoplanet detection in the presence of experimental imperfections

arXiv

引用

arXiv 2025年

作者： Linowski, Tomasz Schlichtholz, Konrad Sorelli, Giacomo International Centre for Theory of Quantum Technologies University of Gdansk Gdańsk80-308 Poland Fraunhofer IOSB Ettlingen Fraunhofer Institute of Optronics System Technologies and Image Exploitation Gutleuthausstr. 1 Ettlingen76275 Germany

Ideal spatial demultiplexing (SPADE) is proven to be a quantum-optimal tool for exoplanet detection, i.e., asymmetric source discrimination. However, recent investigations into the related problems of separation estimation and symmetric source discrimination showed its efficiency to be limited in the presence of noise. In this work, we use analytical tools to scrutinize the practical applicability of SPADE and derive the associated optimal decision strategy for exoplanet detection in the presence of experimental imperfections. On the one hand, we find that the probability of detection of noisy SPADE has the same scaling with planet-star separation and relative brightness as conventional techniques, such as direct imaging and coronagraphs. On the other hand, we prove that, due to a superior scaling coefficient under realistic noise conditions, SPADE remains the most efficient method for practical exoplanet detection in the sub-Rayleigh regime. Copyright © 2025, The Authors. All rights reserved.

关键词： Optimal detection

来源：评论

学校读者我要写书评

暂无评论

Generative-Contrastive Learning for Self-Supervised Latent Representations of 3D Shapes from Multi-Modal Euclidean Input

arXiv

引用

arXiv 2023年

作者： Wu, Chengzhi Pfrommer, Julius Zhou, Mingyuan Beyerer, Jürgen Institute for Anthropomatics and Robotics Karlsruhe Institute of Technology Germany Fraunhofer Institute of Optronics System Technologies and Image Exploitation IOSB Germany Fraunhofer Center for Machine Learning Germany

We propose a combined generative and contrastive neural architecture for learning latent representations of 3D volumetric shapes. The architecture uses two encoder branches for voxel grids and multi-view images from the same underlying shape. The main idea is to combine a contrastive loss between the resulting latent representations with an additional reconstruction loss. That helps to avoid collapsing the latent representations as a trivial solution for minimizing the contrastive loss. A novel switching scheme is used to cross-train two encoders with a shared decoder. The switching scheme also enables the stop gradient operation on a random branch. Further classification experiments show that the latent representations learned with our self-supervised method integrate more useful information from the additional input data implicitly, thus leading to better reconstruction and classification performance. Copyright © 2023, The Authors. All rights reserved.

关键词： Classification (of information)

来源：评论

学校读者我要写书评

暂无评论

Best-of-Breed: Service-Oriented Integration of Artificial Intelligence in Interoperable Educational Ecosystems 11th

Best-of-Breed: Service-Oriented Integration of Artificial In...

引用

11th International Workshop on Learning Technology for Education Challenges, LTEC 2023

作者： Krauss, Christopher Streicher, Alexander Poxleitner, Eva Altun, Daniela Mueller, Joanna An, Truong-Sinh Mueller, Christoph Fraunhofer Institute for Open Communication Systems FOKUS Kaiserin-Augusta-Allee 31 Berlin10589 Germany Fraunhofer Institute of Optronics System Technologies and Image Exploitation IOSB Fraunhoferstraße 1 Karlsruhe76131 Germany Fraunhofer Academy Hansastraße 27c Munich80686 Germany Fraunhofer Institute for Communication Information Processing and Ergonomics FKIE Fraunhoferstraße 20 Wachtberg53343 Germany

ISBN: (纸本)9783031347535

Artificial Intelligence (AI) offers great potential for optimizing learning processes, teaching methods, learning content, or organizational procedures. However, the success of AI components in educational environments is by no means guaranteed and depends on several conditions in their respective learning settings. In this article, we analyze requirements that are often addressed prior to introducing AI features. We address organizational, methodological, didactical, content-related, and technical challenges. The research question of this work is how AI features can best be incorporated into modern educational system landscapes to create sustainable system architectures that are accepted and perceived as added value by users. Thereby, the article discusses two approaches to software architecture: Best-of-Suite (for monolithic architectures) and Best-of-Breed (for service-oriented architectures). Monolithic systems offer a wide range of functions, can be offered by a single provider but can become difficult to manage and create dependencies. Specialized and service-oriented systems, in turn, consist of modular functions handled by specialized services, are more flexible and scalable, and can be integrated with a wide range of tools and services, but require more effort to set up and manage. We explain why the Best-of-Breed strategy is a sensible approach to the use of AI components, how this can be implemented sustainably with the help of a middleware component, and we report on the user experiences from a field test. While in this work we evaluate the implemented system with a cybersecurity training as an on-the-job course, the middleware has been successfully used in other educational contexts, as well. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Interoperability

来源：评论

学校读者我要写书评

暂无评论

The GOOSE Dataset for Perception in Unstructured Environments

arXiv

引用

arXiv 2023年

作者： Mortimer, Peter Hagmanns, Raphael Granero, Miguel Luettel, Thorsten Petereit, Janko Wuensche, Hans-Joachim The Institute for Autonomous Systems Technology University of the Bundeswehr Munich Germany The Fraunhofer Institute of Optronics System Technologies and Image Exploitation Karlsruhe Germany Germany

The potential for deploying autonomous systems can be significantly increased by improving the perception and interpretation of the environment. However, the development of deep learning-based techniques for autonomous systems in unstructured outdoor environments poses challenges due to limited data availability for training and testing. To address this gap, we present the German Outdoor and Offroad Dataset (GOOSE), a comprehensive dataset specifically designed for unstructured outdoor environments. The GOOSE dataset incorporates 10 000 labeled pairs of images and point clouds, which are utilized to train a range of state-of-the-art segmentation models on both image and point cloud data. We open source the dataset, along with an ontology for unstructured terrain, as well as dataset standards and guidelines. This initiative aims to establish a common framework, enabling the seamless inclusion of existing datasets and a fast way to enhance the perception capabilities of various robots operating in unstructured environments. The dataset, pre-trained models for offroad perception, and additional documentation can be found at https://***/. Copyright © 2023, The Authors. All rights reserved.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Single-End-Pumped Tm3+:Ho3+-Codoped all-Fiber Laser at 2120 nm

Single-End-Pumped Tm3+:Ho3+-Codoped all-Fiber Laser at 2120 ...

引用

Conference on Lasers and Electro-Optics Europe (CLEO EUROPE)

作者： Patrick Forster Dieter Panitzek Clement Romano Dominik Lorenz Julian Schneider Marc Eichhorn Christelle Kieleck Fraunhofer IOSB (Institute of Optronics System Technologies and Image Exploitation) Ettlingen Germany Institute of Control Systems (IRS) Karlsruhe Institute of Technology Karlsruhe Germany

ISBN: (纸本)9798350345995

Efficient high-power lasers emitting at a wavelength of around 2.12 $\upmu \mathrm{m}$ are of increasing interest for applications taking advantage of a broad atmospheric transmission window. Typically, Ho 3+ -doped fiber lasers (HDFL) are used to address this wavelength region due to much higher emission cross sections compared to their 2 $\upmu \mathrm{m}$ equivalent Thulium lasers [1], [2]. Recently, Tm 3+ :Ho 3+ -codoped fiber lasers were demonstrated with a slope efficiency as high as 56 % [3] and output powers up to 260 W [4], thus making them a sound substitute to HDFLs for operation above 2.1 $\upmu \mathrm{m}$ . The free-space setup shown in [4] is now transferred to a more integrated and rugged all- fiber version. To the best of our knowledge, we show the first realization of a Tm 3+ :Ho 3+ -codoped all-fiber laser emitting at 2120 nm with a slope efficiency close to 50 %.

关键词： Diode lasers Fiber Bragg gratings Fiber lasers High power lasers Lasers Thulium

来源：评论

学校读者我要写书评

暂无评论

Terrestrial Visual Dust Density Estimation Based On Deep Learning

Terrestrial Visual Dust Density Estimation Based On Deep Lea...

引用

IEEE International Symposium on Geoscience and Remote Sensing (IGARSS)

作者： Andreas Michel Martin Weinmann Fabian Schenkel Thomas Gomez Mark Falvey Rainer Schmitz Wolfgang Middelmann Stefan Hinz Fraunhofer Institute of Optronics System Technologies and Image Exploitation IOSB Germany Institute of Photogrammetry and Remote Sensing Karlsruhe Institute of Technology Germany Meteodata Chile

Airborne dust has a broad impact from climate to human health. Extensive dust monitoring can lead to identifying environmental hazards and developing mitigation strategies. However, conventional dust measuring devices are usually expensive and limited for the monitoring of the spatial characteristics of dust. Available RGB camera systems might be a potential tool for the measurement of these spatial characteristics, but the automatic detection of airborne dust within these images is not well-researched. The challenges for the required algorithm for such an automatic detection are manifold, including the opaqueness, the wide range of possible density levels, the visual similarity to effects like smoke or clouds, and the fuzzy boundaries of airborne dust. In order to face these challenges in the underexplored research field of detecting airborne dust in terrestrial RGB images, we propose DeepDust. DeepDust is a dust density estimation neural network and exploits convolutional-based multi-level embeddings to merge features from different resolutions and semantic levels. Due to the absence of existing methods in our research field, we compare results achieved by our DeepDust with techniques from the crowd counting and monocular depth estimation domain on the Meteodata dust dataset. Our DeepDust outperforms the other evaluated approaches regarding regression ability by a wide margin.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A Maritime Pointing and Tracking Model for Free Space Optical Communications in Military Use Cases

A Maritime Pointing and Tracking Model for Free Space Optica...

引用

International Conference on Military Communications and Information systems (ICMCIS)

作者： Daniel Eames Cheng Chen Stefen Videv Douglas McDonald Harald Haas Iain ME Butler Defence Science and Technology Laboratory Portsmouth United Kingdom LiFi Research and Development Centre University of Strathclyde Glasgow United Kingdom Fraunhofer Institute of Optronics System Technologies and Image Exploitation System Technologies and Image Exploitation Karlsruhe Germany

Due to increasing congestion found in the electro-magnetic spectrum, alternatives must be explored for future communication capabilities. Free Space Optical Communications (FSOC) has the potential to achieve secure, reliable, pervasive, and high-performance wireless connectivity. In this paper, the advantages / disadvantages of the application of FSOC in military use cases will be discussed with a review of the related research. We will provide a case study on the simulation of a camera-based pointing and tracking system deployed on an above-water platforms (Queen Elizabeth Class aircraft carrier / Archer Class patrol vessel). This paper was originally presented at the NATO Science and Technology Organization Symposium (ICMCIS) organized by the Information systems Technology (IST) Panel, IST-200RSY - the ICMCIS, held in Skopje, North Macedonia, 16–17 May 2023

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：