检索结果-内蒙古大学图书馆

29th International Conference on Artificial Neural Networks (ICANN)

作者： Marino, Michael Schroeter, Georg Heidemann, Gunther Hertzberg, Joachim Univ Osnabruck Wachsbleiche 27 D-49090 Osnabruck Germany DFKI Lab Niedersachsen Plan Based Robot Control Dept Berghoffstr 11 D-49090 Osnabruck Germany

ISBN: (纸本)9783030616090;9783030616083

We propose a new analysis technique for neural networks, Neurodynamical Agglomerative Analysis (NAA), an analysis pipeline designed to compare class representations within a given neural network model. The proposed pipeline results in a hierarchy of class relationships implied by the network representation, i.e. a semantic hierarchy analogous to a human-made ontological view of the relevant classes. We use networks pretrained on the ImageNet benchmark dataset to infer semantic hierarchies and show the similarity to human-made semantic hierarchies by comparing them with the WordNet ontology. Further, we show using MNIST training experiments that class relationships extracted using NAA appear to be invariant to random weight initializations, tending toward equivalent class relationships across network initializations in sufficiently parameterized networks.

关键词： Neural network theory Deep learning Cognitive models

来源：评论

学校读者我要写书评

暂无评论

Cognitive Weeding: An Approach to Single-plant Specific Weed Regulation

引用

KUNSTLICHE INTELLIGENZ 2023年第2-4期37卷 175-181页

作者： Niemeyer, Mark Renz, Marian Pukrop, Maren Hagemann, David Zurheide, Tim Di Marco, Daniel Hoeferlin, Markus Stark, Philipp Rahe, Florian Igelbrink, Matthias Jenz, Mario Jarmer, Thomas Trautz, Dieter Stiene, Stefan Hertzberg, Joachim German Res Ctr Artificial Intelligence Plan Based Robot Control Grp Osnabruck Germany Osnabruck Univ Osnabruck Germany Univ Appl Sci Osnabruck Osnabruck Germany Farming Revolut Ludwigsburg Germany Amazonen Werke Hasbergen Germany

This paper provides a comprehensive overview of the architecture required to implement selective weeding in arable farming, as developed within the Cognitive Weeding project. This end-to-end architecture begins with data acquisition utilizing drones, robots, or agricultural machinery, followed by data management, AI-based data annotation, knowledge-based inference to determine the necessary treatment, resulting in an application map for selective hoeing. The paper meticulously details the various components of the architecture and illustrates through examples how they are interconnected.

关键词： selective weeding site-specific weed management precision agriculture expert system

来源：评论

学校读者我要写书评

暂无评论

Geometry-based Regularisation for Dense Image Matching Via Uncertainty-Driven Depth Propagation 24

Geometry-Based Regularisation for Dense Image Matching Via U...

引用

2020 24th ISPRS Congress on Technical Commission II

作者： Höllmann, M. Mehltretter, M. Heipke, C. Plan-Based Robot Control Group German Research Center for Artificial Intelligence Osnabrück Germany Institute of Photogrammetry and GeoInformation Leibniz University Hannover Germany

In the present work, an uncertainty-driven geometry-based regularisation for the task of dense stereo matching is presented. The objective of the regularisation is the reduction of ambiguities in the depth reconstruction process, which exist due to the ill-posed nature of this task. based on cost and uncertainty information computed beforehand, pixels are selected, whose depth information can be determined correctly with a high probability. This depth information assumed to be of high confidence is initially used to construct a triangle mesh, which is interpreted as surface approximation of the imaged scene and allows to propagate the confident depth information of the triangle vertices within local neighbourhoods. The proposed method further computes confidence scores for propagated depth estimates, which are used to fuse this depth information with the previously computed cost information, introducing a regularisation into the data term of global optimisation methods. Furthermore, based on the propagated depth information the local smoothness assumption of global optimisation methods is adjusted. Instead of fronto-parallel planes, the method presumes planes, which are parallel to the propagated depth information. The performance of the proposed regularisation approach is evaluated in combination with a global optimisation method. For a quantitative and qualitative evaluation two commonly employed and well-established stereo datasets are used. The proposed method shows significant improvements in accuracy on both datasets and for two different cost computation methods. Especially in unstructured areas, artefacts in the disparity maps are reduced. © 2020 Copernicus GmbH. All rights reserved.

关键词： Image matching

来源：评论

学校读者我要写书评

暂无评论

Hyperspectral 3D Point Cloud Segmentation Using RandLA-Net 17th

Hyperspectral 3D Point Cloud Segmentation Using RandLA-Net

引用

17th International Conference on Intelligent Autonomous Systems (IAS)

作者： Mitschke, Isaak Wiemann, Thomas Igelbrink, Felix Hertzberg, Joachim Osnabruck Univ Inst Comp Sci Berghoffstr 11 D-49090 Osnabruck Germany DFKI Niedersachsen Lab Plan Based Robot Control Berghoffstr 11 D-49090 Osnabruck Germany

ISBN: (纸本)9783031222153;9783031222160

Point clouds are commonly used in robotics to represent 3D maps. To gain further understanding of their content, it is useful to annotate such maps semantically. To segment 3D point clouds with RGB values, different solutions exist. In machine learning, pre-trained classifiers are used for this purpose. Since it is not always possible to differentiate between entities relying solely on RGB information, hyperspectral histograms can augment the 3D data. The aim of this work is to evaluate, if hyperspectral information can improve the segmentation results for ambiguous objects, e.g., streets, sidewalks, and cars using established deep learning methods. Given the reported performance on geometrical 3D data and the possibility to directly integrate point annotations, we extended the neural network RandLA-Net. In addition to the evaluation of RandLA-Net in this context, we also provide a reference dataset consisting of semantically annotated hyperspectral 3D point clouds.

关键词： Point clouds Semantic segmentation Deep learning Hyperspectral imaging

来源：评论

学校读者我要写书评

暂无评论

AI-based Maize and Weeds Detection on the Edge with CornWeed Dataset 18

AI-Based Maize and Weeds Detection on the Edge with CornWeed...

引用

18th Conference on Computer Science and Intelligence Systems, FedCSIS 2023

作者： Iqbal, Naeem Manss, Christoph Scholz, Christian Konig, Daniel Igelbrink, Matthias Ruckelshausen, Arno Dfki Plan-based Robot Control Osnabrueck Germany Dfki Marine Perception Oldenburg Germany University of Applied Sciences Osnabrueck Faculty of Engineering and Computer Science Osnabrueck Germany

ISBN: (纸本)9788396744784

Artificial intelligence (AI) is used more heavily in agricultural applications. Yet, the lack of wireless-fidelity (Wi-Fi) connections on agricultural fields makes AI cloud services unavailable. Consequently, AI models have to be processed directly on the edge. In this paper, we evaluate state-of-the-art detection algorithms for their use in agriculture, in particular plant detection. Thus, this paper presents the CornWeed data set, which has been recorded on farm machines, showing labelled maize crops and weeds for plant detection. The paper provides accuracies for the state-of-the-art detection algorithms on the CornWeed data set, as well as frames per second (FPS) metrics for the considered networks on multiple edge devices. Moreover, for the FPS analysis, the detection algorithms are converted to open neural network exchange (ONNX) and TensoRT engine files as they could be used as future standards for model exchange. © 2023 Polish Information Processing Society.

关键词： Data acquisition

来源：评论

学校读者我要写书评

暂无评论

Rmagine: 3D Range Sensor Simulation in Polygonal Maps via Ray Tracing for Embedded Hardware on Mobile robots

Rmagine: 3D Range Sensor Simulation in Polygonal Maps via Ra...

引用

IEEE International Conference on robotics and Automation (ICRA)

作者： Mock, Alexander Wiemann, Thomas Hertzberg, Joachim Osnabruck Univ Inst Comp Sci Knowledge Based Syst Grp Berghoffstr 11 D-49090 Osnabruck Germany Osnabruck Univ Inst Comp Sci Amonomous Robot Grp Berghoffstr 11 D-49090 Osnabruck Germany DFKI Niedersachsen Plan Based Robot Control Berghoffstr 11 D-49090 Osnabruck Germany

ISBN: (纸本)9798350323658

Sensor simulation has emerged as a promising and powerful technique to find solutions to many real-world robotic tasks like localization and pose tracking. However, commonly used simulators have high hardware requirements and are therefore used mostly on high-end computers. In this paper, we present an approach to simulate range sensors directly on embedded hardware of mobile robots that use triangle meshes as environment map. This library, called Rmagine, allows a robot to simulate sensor data for arbitrary range sensors directly on board via ray tracing. Since robots typically only have limited computational resources, Rmagine aims at being flexible and lightweight, while scaling well even to large environment maps. It runs on several platforms like Laptops or embedded computing boards like NVIDIA Jetson by putting an unified API over the specific proprietary libraries provided by the hardware manufacturers. This work is designed to support the future development of robotic applications depending on simulation of range data that could previously not be computed in reasonable time on mobile systems.

关键词： Ray tracing

来源：评论

学校读者我要写书评

暂无评论

PHEATPRUNER: Interpretable Data-centric Feature Selection for Multivariate Time Series Classification through Persistent Homology

arXiv

引用

arXiv 2025年

作者： Pham, Anh-Duy Kashongwe, Olivier Basole Atzmueller, Martin Römer, Tim Osnabrück University Joint Lab Artificial Intelligence & Data Science Osnabrück Germany Osnabrück University Institute of Mathematics Osnabrück Germany Osnabrück University Semantic Information Systems Group Osnabrück Germany Research Department Plan-Based Robot Control Osnabrück Germany Department of Sensors and Modelling Potsdam Germany

Balancing performance and interpretability in multivariate time series classification is a significant challenge due to data complexity and high dimensionality. This paper introduces PHEATPRUNER, a method integrating persistent homology and sheaf theory to address these challenges. Persistent homology facilitates the pruning of up to 45% of the applied variables while maintaining or enhancing the accuracy of models such as Random Forest, CatBoost, XGBoost, and LightGBM, all without depending on posterior probabilities or supervised optimization algorithms. Concurrently, sheaf theory contributes explanatory vectors that provide deeper insights into the data’s structural nuances. The approach was validated using the UEA Archive and a mastitis detection dataset for dairy cows. The results demonstrate that PHEATPRUNER effectively preserves model accuracy. Furthermore, our results highlight PHEATPRUNER’s key features, i. e., simplifying complex data and offering actionable insights without increasing processing time or complexity. This method bridges the gap between complexity reduction and interpretability, suggesting promising applications in various fields. Copyright © 2025, The Authors. All rights reserved.

关键词： Data reduction

来源：评论

学校读者我要写书评

暂无评论

MICP-L: Mesh-based ICP for robot Localization Using Hardware-Accelerated Ray Casting

MICP-L: Mesh-based ICP for Robot Localization Using Hardware...

引用

2024 International Conference on Intelligent robots and Systems

作者： Mock, Alexander Wiemann, Thomas Puetz, Sebastian Hertzberg, Joachim Osnabruck Univ Inst Comp Sci Knowledge Based Syst Grp Hamburger Str 24 D-49084 Osnabruck Germany Fulda Univ Appl Sci Robot Comp Sci Fulda Germany Nat Robots GmbH Osnabruck Germany DFKI Plan Based Robot Control Hamburger Str 24 D-49084 Osnabruck Germany

ISBN: (纸本)9798350377712;9798350377705

Triangle mesh maps are a versatile 3D environment representation for robots to navigate in challenging indoor and outdoor environments exhibiting tunnels, hills and varying slopes. To make use of these mesh maps, methods are needed to accurately localize robots in such maps to perform essential tasks like path planning and navigation. We present Mesh ICP Localization (MICP-L), a novel and computationally efficient method for registering one or more range sensors to a triangle mesh map to continuously localize a robot in 6D, even in GPS-denied environments. We accelerate the computation of ray casting correspondences (RCC) between range sensors and mesh maps by supporting different parallel computing devices like multicore CPUs, GPUs and the latest NVIDIA RTX hardware. By additionally transforming the covariance computation into a reduction operation, we can optimize the initial guessed poses in parallel on CPUs or GPUs, making our implementation applicable in real-time on many architectures. We demonstrate the robustness of our localization approach with datasets from agricultural, aerial, and automotive domains.

关键词： Mesh generation

来源：评论

学校读者我要写书评

暂无评论

Towards model-based automation of plant-specific weed regulation 43

Towards model-based automation of plant-specific weed regula...

引用

43. Jahrestagung der Gesellschaft fur Informatik in der Land-, Forst- und Ernahrungswirtschaft - Resiliente Agri-Food-Systeme: Herausforderungen und Losungsansatze, GIL 2023 - 43rd Annual Conference of the Society for Informatics in Agriculture, Forestry, and Food Industry - Resilient Agri-Food Systems: Challenges and Solutions, GIL 2023

作者： Renz, Marian Niemeyer, Mark Hertzberg, Joachim German Research Center for Artificial Intelligence Plan-Based Robot Control Group Berghoffstr. 11 Osnabrück49090 Germany Osnabrück University Knowledge-Based Systems Group Wachsbleiche 27 Osnabrück49090 Germany

ISBN: (纸本)9783885797241

Weeds are commonly known as a major factor for yield losses in agriculture, competing with crops for resources like nutrients, water, and light. However, keeping specific weeds could benefit agricultural sites for example by nitrogen fixation, erosion protection, or increasing biodiversity. This comes with technological challenges like plant detection and classification, damage estimation, and selective removal. This paper presents a model-based approach to the problem of damage estimation of perceived plants. The system uses contextual and background knowledge in the form of rules about the plant count per square meter and the distance to the nearest crop together with thresholds for each weed species. The functionality is demonstrated using an artificial dataset and exemplary thresholds, showing the potential of using knowledge about plant-crop interactions for more sophisticated weed control systems. © 2023 Gesellschaft fur Informatik (GI). All rights reserved.

关键词： Crops

来源：评论

学校读者我要写书评

暂无评论

Leveraging Explainable AI Methods Towards Identifying Classification Issues on IDS Datasets 48

Leveraging Explainable AI Methods Towards Identifying Classi...

引用

48th Conference on Local Computer Networks

作者： Lanfer, Eric Sylvester, Sophia Aschenbruck, Nils Atzmueller, Martin Osnabruck Univ Inst Comp Sci Osnabruck Germany Norwegian Univ Sci & Technol NTNU Dept Mental Hlth Trondheim Norway German Res Ctr AI DFKI Plan Based Robot Control Osnabruck Germany

ISBN: (纸本)9798350300741;9798350300734

Nowadays, anomaly-based network intrusion detection system (NIDS) still have limited real-world applications;this is particularly due to false alarms, a lack of datasets, and a lack of confidence. In this paper, we propose to use explainable artificial intelligence (XAI) methods for tackling these issues. In our experimentation, we train a random forest (RF) model on the NSL-KDD dataset, and use SHAP to generate global explanations. We find that these explanations deviate substantially from domain expertise. To shed light on the potential causes, we analyze the structural composition of the attack classes. There, we observe severe imbalances in the number of records per attack type subsumed in the attack classes of the NSL-KDD dataset, which could lead to generalization and overfitting regarding classification. Hence, we train a new RF classifier and SHAP explainer directly on the attack types. Classification performance is considerably improved, and the new explanations are matching the expectations based on domain knowledge better. Thus, we conclude that the imbalances in the dataset bias classification and consequently also the results of XAI methods like SHAP. However, the XAI methods can also be employed to find and debug issues and biases in the data and the applied model. Furthermore, the debugging results in higher trustworthiness of anomaly-based NIDS.

关键词： Explainable AI Cybersecurity Network Intrusion Detection NSL-KDD Feature Importance SHAP

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：