PUMA publications for /https://puma.uni-kassel.de/PUMA RSS feed for /2024-03-19T05:21:33+01:00Graph Pooling Provably Improves Expressivityhttps://puma.uni-kassel.de/bibtex/2e1ea17f1ffa6ae37f7e79d245f6b5d6b/04068750040687502024-02-21T17:21:46+01:00imported itegpub isac-www <span class="authorEditorList"><span itemtype="http://schema.org/Person" itemscope="itemscope" itemprop="author"><a title="Veronica Lachi" itemprop="url" href="/author/Veronica%20Lachi"><span itemprop="name">V. Lachi</span></a></span>, <span itemtype="http://schema.org/Person" itemscope="itemscope" itemprop="author"><a title="Alice Moallemy-Oureh" itemprop="url" href="/author/Alice%20Moallemy-Oureh"><span itemprop="name">A. Moallemy-Oureh</span></a></span>, <span itemtype="http://schema.org/Person" itemscope="itemscope" itemprop="author"><a title="Andreas Roth" itemprop="url" href="/author/Andreas%20Roth"><span itemprop="name">A. Roth</span></a></span>, und <span itemtype="http://schema.org/Person" itemscope="itemscope" itemprop="author"><a title="Pascal Welke" itemprop="url" href="/author/Pascal%20Welke"><span itemprop="name">P. Welke</span></a></span>. </span><span itemtype="http://schema.org/Book" itemscope="itemscope" itemprop="isPartOf"><em><span itemprop="name">Workshop on New Frontiers in Graph Learning, NeurIPS</span>, </em></span>(<em><span>2023<meta content="2023" itemprop="datePublished"/></span></em>)Wed Feb 21 17:21:46 CET 2024Workshop on New Frontiers in Graph Learning, NeurIPSGraph Pooling Provably Improves Expressivity2023imported itegpub isac-www In the domain of graph neural networks (GNNs), pooling operators are fundamental to reduce the size of the graph by simplifying graph structures and vertex features. Recent advances have shown that well-designed pooling operators, coupled with message-passing layers, can endow hierarchical GNNs with an expressive power regarding the graph isomorphism test that is equal to the Weisfeiler-Leman test. However, the ability of hierarchical GNNs to increase expressive power by utilizing graph coarsening was not yet explored. This results in uncertainties about the benefits of pooling operators and a lack of sufficient properties to guide their design. In this work, we identify conditions for pooling operators to generate WL-distinguishable coarsened graphs from originally WL-indistinguishable but non-isomorphic graphs. Our conditions are versatile and can be tailored to specific tasks and data characteristics, offering a promising avenue for further research.Spatio-Temporal Attention Graph Neural Network for Remaining Useful Life Predictionhttps://puma.uni-kassel.de/bibtex/2ad927931297c4668fb4cda3fa215b717/04068750040687502024-02-21T17:09:30+01:00imported itegpub isac-www Spatio-Temporal_Attention Remaining_Useful_Life Graph_Neural_Network RUL_Prediction Clustering_Normalization <span class="authorEditorList"><span itemtype="http://schema.org/Person" itemscope="itemscope" itemprop="author"><a title="Zhixin Huang" itemprop="url" href="/author/Zhixin%20Huang"><span itemprop="name">Z. Huang</span></a></span>, <span itemtype="http://schema.org/Person" itemscope="itemscope" itemprop="author"><a title="Yujiang He" itemprop="url" href="/author/Yujiang%20He"><span itemprop="name">Y. He</span></a></span>, und <span itemtype="http://schema.org/Person" itemscope="itemscope" itemprop="author"><a title="Bernhard Sick" itemprop="url" href="/author/Bernhard%20Sick"><span itemprop="name">B. Sick</span></a></span>. </span><span itemtype="http://schema.org/Book" itemscope="itemscope" itemprop="isPartOf"><em><span itemprop="name">Computational Science and Computational Intelligence (CSCI)</span>, </em></span><em><span itemprop="publisher">IEEE</span>, </em>(<em><span>2023<meta content="2023" itemprop="datePublished"/></span></em>)<em>(accepted).</em>Wed Feb 21 17:09:30 CET 2024Computational Science and Computational Intelligence (CSCI)(accepted)Spatio-Temporal Attention Graph Neural Network for Remaining Useful Life Prediction2023imported itegpub isac-www Spatio-Temporal_Attention Remaining_Useful_Life Graph_Neural_Network RUL_Prediction Clustering_Normalization RUL prediction plays a crucial role in the health management of industrial systems. Given the increasing complexity of systems, data-driven predictive models have attracted significant research interest. Upon reviewing the existing literature, it appears that many studies either do not fully integrate both spatial and temporal features or employ only a single attention mechanism. Furthermore, there seems to be inconsistency in the choice of data normalization methods, particularly concerning operating conditions, which might influence predictive performance. To bridge these observations, this study presents the Spatio-Temporal Attention Graph Neural Network. Our model combines graph neural networks and temporal convolutional neural networks for spatial and temporal feature extraction, respectively. The cascade of these extractors, combined with multihead attention mechanisms for both spatio-temporal dimensions, aims to improve predictive precision and refine model explainability. Comprehensive experiments were conducted on the CMAPSS dataset to evaluate the impact of unified versus clustering normalization. The findings suggest that our model performs state-of-the-art results using only the unified normalization. Additionally, when dealing with datasets with multiple operating conditions, cluster normalization enhances the performance of our proposed model by up to 27%.Organic Computing -- Doctoral Dissertation Colloquium 2021https://puma.uni-kassel.de/bibtex/27647b51cffa0dc81c1bdc564a4929c52/04068750040687502024-02-21T16:43:48+01:00imported itegpub isac-www <span class="authorEditorList"><span itemtype="http://schema.org/Person" itemscope="itemscope" itemprop="editor"><a title="Sven Tomforde" itemprop="url" href="/author/Sven%20Tomforde"><span itemprop="name">S. Tomforde</span></a></span>, und <span itemtype="http://schema.org/Person" itemscope="itemscope" itemprop="editor"><a title="Christian Krupitzer" itemprop="url" href="/author/Christian%20Krupitzer"><span itemprop="name">C. Krupitzer</span></a></span> (Hrsg.).
. </span><em>Intelligent Embedded Systems </em><em><span itemprop="publisher">kassel university press</span>, </em>(<em><span>2022<meta content="2022" itemprop="datePublished"/></span></em>)Wed Feb 21 16:43:48 CET 2024Intelligent Embedded SystemsOrganic Computing -- Doctoral Dissertation Colloquium 2021202022imported itegpub isac-www Height Change Feature Based Free Space Detectionhttps://puma.uni-kassel.de/bibtex/2c68e47214fba4ebbe449b410278ca949/04068750040687502024-02-21T16:06:59+01:00imported itegpub isac-www <span class="authorEditorList"><span itemtype="http://schema.org/Person" itemscope="itemscope" itemprop="author"><a title="Steven Schreck" itemprop="url" href="/author/Steven%20Schreck"><span itemprop="name">S. Schreck</span></a></span>, <span itemtype="http://schema.org/Person" itemscope="itemscope" itemprop="author"><a title="Hannes Reichert" itemprop="url" href="/author/Hannes%20Reichert"><span itemprop="name">H. Reichert</span></a></span>, <span itemtype="http://schema.org/Person" itemscope="itemscope" itemprop="author"><a title="Manuel Hetzel" itemprop="url" href="/author/Manuel%20Hetzel"><span itemprop="name">M. Hetzel</span></a></span>, <span itemtype="http://schema.org/Person" itemscope="itemscope" itemprop="author"><a title="Konrad Doll" itemprop="url" href="/author/Konrad%20Doll"><span itemprop="name">K. Doll</span></a></span>, und <span itemtype="http://schema.org/Person" itemscope="itemscope" itemprop="author"><a title="Bernhard Sick" itemprop="url" href="/author/Bernhard%20Sick"><span itemprop="name">B. Sick</span></a></span>. </span><span itemtype="http://schema.org/Book" itemscope="itemscope" itemprop="isPartOf"><em><span itemprop="name">International Conference on Control, Mechatronics and Automation (ICCMA)</span>, </em></span><em>Seite <span itemprop="pagination">171--176</span>. </em><em><span itemprop="publisher">IEEE</span>, </em>(<em><span>2023<meta content="2023" itemprop="datePublished"/></span></em>)Wed Feb 21 16:06:59 CET 2024International Conference on Control, Mechatronics and Automation (ICCMA)171--176Height Change Feature Based Free Space Detection2023imported itegpub isac-www We present a novel method for free space detection in dynamic environments like factory sites, crucial for autonomous forklifts to avoid collisions. It introduces a technique for fast surface normal estimation using spherical projected LiDAR data, which helps in detecting free space efficiently and in real-time. The method's effectiveness is proven with a 50.90% mIoU score on the Semantic KITTI dataset at 105 Hz, and a 63.30% mIoU score on a factory site dataset at 54 Hz.What Does Really Count? Estimating Relevance of Corner Cases for Semantic Segmentation in Automated Drivinghttps://puma.uni-kassel.de/bibtex/2dd6f5a976f1b9423ffe3a36248745b42/04068750040687502024-02-21T15:35:57+01:00imported itegpub isac-www <span class="authorEditorList"><span itemtype="http://schema.org/Person" itemscope="itemscope" itemprop="author"><a title="Jasmin Breitenstein" itemprop="url" href="/author/Jasmin%20Breitenstein"><span itemprop="name">J. Breitenstein</span></a></span>, <span itemtype="http://schema.org/Person" itemscope="itemscope" itemprop="author"><a title="Florian Heidecker" itemprop="url" href="/author/Florian%20Heidecker"><span itemprop="name">F. Heidecker</span></a></span>, <span itemtype="http://schema.org/Person" itemscope="itemscope" itemprop="author"><a title="Maria Lyssenko" itemprop="url" href="/author/Maria%20Lyssenko"><span itemprop="name">M. Lyssenko</span></a></span>, <span itemtype="http://schema.org/Person" itemscope="itemscope" itemprop="author"><a title="Daniel Bogdoll" itemprop="url" href="/author/Daniel%20Bogdoll"><span itemprop="name">D. Bogdoll</span></a></span>, <span itemtype="http://schema.org/Person" itemscope="itemscope" itemprop="author"><a title="Maarten Bieshaar" itemprop="url" href="/author/Maarten%20Bieshaar"><span itemprop="name">M. Bieshaar</span></a></span>, <span itemtype="http://schema.org/Person" itemscope="itemscope" itemprop="author"><a title="J. Marius Zöllner" itemprop="url" href="/author/J.%20Marius%20Z%c3%b6llner"><span itemprop="name">J. Zöllner</span></a></span>, <span itemtype="http://schema.org/Person" itemscope="itemscope" itemprop="author"><a title="Bernhard Sick" itemprop="url" href="/author/Bernhard%20Sick"><span itemprop="name">B. Sick</span></a></span>, und <span itemtype="http://schema.org/Person" itemscope="itemscope" itemprop="author"><a title="Tim Fingscheidt" itemprop="url" href="/author/Tim%20Fingscheidt"><span itemprop="name">T. Fingscheidt</span></a></span>. </span><span itemtype="http://schema.org/Book" itemscope="itemscope" itemprop="isPartOf"><em><span itemprop="name">Workshop on roBustness and Reliability of Autonomous Vehicles in the Open-world (BRAVO), ICCV</span>, </em></span><em>Seite <span itemprop="pagination">3991--4000</span>. </em>(<em><span>2023<meta content="2023" itemprop="datePublished"/></span></em>)Wed Feb 21 15:35:57 CET 2024Workshop on roBustness and Reliability of Autonomous Vehicles in the Open-world (BRAVO), ICCV3991--4000What Does Really Count? Estimating Relevance of Corner Cases for Semantic Segmentation in Automated Driving2023imported itegpub isac-www In safety-critical applications such as automated driving, perception errors may create an imminent risk to vulnerable road users (VRU). To mitigate the occurrence of unexpected and potentially dangerous situations, so-called corner cases, perception models are trained on a huge amount of data. However, the models are typically evaluated using task-agnostic metrics, which do not reflect the severity of safety-critical misdetections. Consequently, misdetections with particular relevance for the safe driving task should entail a more severe penalty during evaluation to pinpoint corner cases in large-scale datasets. In this work, we propose a novel metric IoUw that exploits relevance on the pixel level of the semantic segmentation output to extend the notion of the intersection over union (IoU) by emphasizing small areas of an image affected by corner cases. We (i) employ IoUw to measure the effect of pre-defined relevance criteria on the segmentation evaluation, and (ii) use the relevance-adapted IoUw to refine the identification of corner cases. In our experiments, we investigate vision-based relevance criteria and physical attributes as per-pixel criticality to factor in the imminent risk, showing that IoUw precisely accentuates the relevance of corner cases.