Jochen Garcke

Lokesh Veeramacheneni, Moritz Wolter, Reinhard Klein et al.

We introduce canonical weight normalization for convolutional neural networks. Inspired by the canonical tensor decomposition, we express the weight tensors in so-called canonical networks as scaled sums of outer vector products. In particular, we train network weights in the decomposed form, where scale weights are optimized separately for each mode. Additionally, similarly to weight normalization, we include a global scaling parameter. We study the initialization of the canonical form by running the power method and by drawing randomly from Gaussian or uniform distributions. Our results indicate that we can replace the power method with cheaper initializations drawn from standard distributions. The canonical re-parametrization leads to competitive normalization performance on the MNIST, CIFAR10, and SVHN data sets. Moreover, the formulation simplifies network compression. Once training has converged, the canonical form allows convenient model-compression by truncating the parameter sums.

Rodrigo Iza-Teran, Jochen Garcke

We propose a new data analysis approach for the efficient post-processing of bundles of finite element data from numerical simulations. The approach is based on the mathematical principles of symmetry. We consider the case where simulations of an industrial product are contained in the space of surface meshes embedded in R^3. Furthermore, we assume that distance preserving transformations exist, albeit unknown, which map simulation to simulation. In this setting, a discrete Laplace-Beltrami operator can be constructed on the mesh, which is invariant to isometric transformations and therefore valid for all simulations. The eigenfunctions of such an operator are used as a common basis for all (isometric) simulations. One can use the projection coefficients instead of the full simulations for further analysis. To extend the idea of invariance, we employ a discrete Fokker-Planck operator, that in the continuous limit converges to an operator invariant to a nonlinear transformation, and use its eigendecomposition accordingly. The data analysis approach is applied to time-dependent datasets from numerical car crash simulations. One observes that only a few spectral coefficients are necessary to describe the data variability, and low dimensional structures are obtained. The eigenvectors are seen to recover different independent variation modes such as translation, rotation, or global and local deformations. An effective analysis of the data from bundles of numerical simulations is made possible, in particular an analysis for many simulations in time.

Sara Hahner, Souhaib Attaiki, Jochen Garcke et al.

We introduce a novel learning-based method for encoding and manipulating 3D surface meshes. Our method is specifically designed to create an interpretable embedding space for deformable shape collections. Unlike previous 3D mesh autoencoders that require meshes to be in a 1-to-1 correspondence, our approach is trained on diverse meshes in an unsupervised manner. Central to our method is a spectral pooling technique that establishes a universal latent space, breaking free from traditional constraints of mesh connectivity and shape categories. The entire process consists of two stages. In the first stage, we employ the functional map paradigm to extract point-to-point (p2p) maps between a collection of shapes in an unsupervised manner. These p2p maps are then utilized to construct a common latent space, which ensures straightforward interpretation and independence from mesh connectivity and shape category. Through extensive experiments, we demonstrate that our method achieves excellent reconstructions and produces more realistic and smoother interpolations than baseline approaches.

Niklas Breustedt, Paolo Climaco, Jochen Garcke et al.

Machine learning techniques paired with the availability of massive datasets dramatically enhance our ability to explore the chemical compound space by providing fast and accurate predictions of molecular properties. However, learning on large datasets is strongly limited by the availability of computational resources and can be infeasible in some scenarios. Moreover, the instances in the datasets may not yet be labelled and generating the labels can be costly, as in the case of quantum chemistry computations. Thus, there is a need to select small training subsets from large pools of unlabelled data points and to develop reliable ML methods that can effectively learn from small training sets. This work focuses on predicting the molecules atomization energy in the QM9 dataset. We investigate the advantages of employing domain knowledge-based data sampling methods for an efficient training set selection combined with informed ML techniques. In particular, we show how maximizing molecular diversity in the training set selection process increases the robustness of linear and nonlinear regression techniques such as kernel methods and graph neural networks. We also check the reliability of the predictions made by the graph neural network with a model-agnostic explainer based on the rate distortion explanation framework.

Barbara Fuchs, Jochen Garcke

Most approximation methods in high dimensions exploit smoothness of the function being approximated. These methods provide poor convergence results for non-smooth functions with kinks. For example, such kinks can arise in the uncertainty quantification of quantities of interest for gas networks. This is due to the regulation of the gas flow, pressure, or temperature. But, one can exploit that for each sample in the parameter space it is known if a regulator was active or not, which can be obtained from the result of the corresponding numerical solution. This information can be exploited in a stochastic collocation method. We approximate the function separately on each smooth region by polynomial interpolation and obtain an approximation to the kink. Note that we do not need information about the exact location of kinks, but only an indicator assigning each sample point to its smooth region. We obtain a global order of convergence of $(p+1)/d$, where $p$ is the degree of the employed polynomials and $d$ the dimension of the parameter space.

Sara Hahner, Felix Kerkhoff, Jochen Garcke

The underlying dynamics and patterns of 3D surface meshes deforming over time can be discovered by unsupervised learning, especially autoencoders, which calculate low-dimensional embeddings of the surfaces. To study the deformation patterns of unseen shapes by transfer learning, we want to train an autoencoder that can analyze new surface meshes without training a new network. Here, most state-of-the-art autoencoders cannot handle meshes of different connectivity and therefore have limited to no generalization capacities to new meshes. Also, reconstruction errors strongly increase in comparison to the errors for the training shapes. To address this, we propose a novel spectral CoSMA (Convolutional Semi-Regular Mesh Autoencoder) network. This patch-based approach is combined with a surface-aware training. It reconstructs surfaces not presented during training and generalizes the deformation behavior of the surfaces' patches. The novel approach reconstructs unseen meshes from different datasets in superior quality compared to state-of-the-art autoencoders that have been trained on these shapes. Our transfer learning errors on unseen shapes are 40% lower than those from models learned directly on the data. Furthermore, baseline autoencoders detect deformation patterns of unseen mesh sequences only for the whole shape. In contrast, due to the employed regional patches and stable reconstruction quality, we can localize where on the surfaces these deformation patterns manifest.

Anahita Pakiman, Jochen Garcke

We consider graph modeling for a knowledge graph for vehicle development, with a focus on crash safety. An organized schema that incorporates information from various structured and unstructured data sources is provided, which includes relevant concepts within the domain. In particular, we propose semantics for crash computer aided engineering (CAE) data, which enables searchability, filtering, recommendation, and prediction for crash CAE data during the development process. This graph modeling considers the CAE data in the context of the R\&D development process and vehicle safety. Consequently, we connect CAE data to the protocols that are used to assess vehicle safety performances. The R\&D process includes CAD engineering and safety attributes, with a focus on multidisciplinary problem-solving. We describe previous efforts in graph modeling in comparison to our proposal, discuss its strengths and limitations, and identify areas for future work.

Haoqing Wu, Alexa Nawotki, Jochen Garcke

Point clouds are a fundamental 3D representation in computer vision, enabling a wide range of perception tasks. However, real-world point clouds often suffer from degradations such as incompleteness, noise, outliers, and irregular density, caused by sensor limitations or occlusions. Recovering clean and detailed shapes from such degraded data is crucial for downstream applications. While existing learning-based methods achieve progress on individual tasks like completion or denoising, they typically rely on global bottleneck features, which lose fine-grained geometry and remain sensitive to varying input quality. We propose a unified 3D restoration network that directly takes point clouds as input and adaptively reconstructs high-quality geometry under diverse degradation scenarios. At the core of our approach is a Pseudo-Query module, implemented within a Transformer backbone, which reformulates geometric translation into two cooperative stages to enhance structural clarity, robustness, and local detail preservation. Extensive experiments on curated benchmarks demonstrate that our approach surpasses state-of-the-art performance in general 3D restoration. It effectively handles complex combinations of completion, deformation, and denoising degradations. With this work, we provide a novel unified, point-only backbone for robust 3D restoration, enabling more versatile 3D perception.

Paolo Climaco, Jochen Garcke

For regression tasks one often leverages large datasets for training predictive machine learning models. However, using large datasets may not be feasible due to computational limitations or high data labelling costs. Therefore, suitably selecting small training sets from large pools of unlabelled data points is essential to maximize model performance while maintaining efficiency. In this work, we study Farthest Point Sampling (FPS), a data selection approach that aims to minimize the fill distance of the selected set. We derive an upper bound for the maximum expected prediction error, conditional to the location of the unlabelled data points, that linearly depends on the training set fill distance. For empirical validation, we perform experiments using two regression models on three datasets. We empirically show that selecting a training set by aiming to minimize the fill distance, thereby minimizing our derived bound, significantly reduces the maximum prediction error of various regression models, outperforming alternative sampling approaches by a large margin. Furthermore, we show that selecting training sets with the FPS can also increase model stability for the specific case of Gaussian kernel regression approaches.

Anahita Pakiman, Jochen Garcke, Axel Schumacher

In this work, we establish a method for abstracting information from Computer Aided Engineering (CAE) into graphs. Such graph representations of CAE data can improve design guidelines and support recommendation systems by enabling the comparison of simulations, highlighting unexplored experimental designs, and correlating different designs. We focus on the load-path in crashworthiness analysis, a complex sub-discipline in vehicle design. The load-path is the sequence of parts that absorb most of the energy caused by the impact. To detect the load-path, we generate a directed weighted graph from the CAE data. The vertices represent the vehicle's parts, and the edges are an abstraction of the connectivity of the parts. The edge direction follows the temporal occurrence of the collision, where the edge weights reflect aspects of the energy absorption. We introduce and assess three methods for graph extraction and an additional method for further updating each graph with the sequences of absorption. Based on longest-path calculations, we introduce an automated detection of the load-path, which we analyse for the different graph extraction methods and weights. Finally, we show how our method for the detection of load-paths helps in the classification and labelling of CAE simulations.

Paolo Climaco, Jochen Garcke

We focus on training machine learning regression models in scenarios where the availability of labeled training data is limited due to computational constraints or high labeling costs. Thus, selecting suitable training sets from unlabeled data is essential for balancing performance and efficiency. For the selection of the training data, we focus on passive and model-agnostic sampling methods that only consider the data feature representations. We derive an upper bound for the expected prediction error of Lipschitz continuous regression models that linearly depends on the weighted fill distance of the training set, a quantity we can estimate simply by considering the data features. We introduce "Density-Aware Farthest Point Sampling" (DA-FPS), a novel sampling method. We prove that DA-FPS provides approximate minimizers for a data-driven estimation of the weighted fill distance, thereby aiming at minimizing our derived bound. We conduct experiments using two regression models across three datasets. The results demonstrate that DA-FPS significantly reduces the mean absolute prediction error compared to other sampling strategies.

Sara Hahner, Jochen Garcke

The analysis of deforming 3D surface meshes is accelerated by autoencoders since the low-dimensional embeddings can be used to visualize underlying dynamics. But, state-of-the-art mesh convolutional autoencoders require a fixed connectivity of all input meshes handled by the autoencoder. This is due to either the use of spectral convolutional layers or mesh dependent pooling operations. Therefore, the types of datasets that one can study are limited and the learned knowledge cannot be transferred to other datasets that exhibit similar behavior. To address this, we transform the discretization of the surfaces to semi-regular meshes that have a locally regular connectivity and whose meshing is hierarchical. This allows us to apply the same spatial convolutional filters to the local neighborhoods and to define a pooling operator that can be applied to every semi-regular mesh. We apply the same mesh autoencoder to different datasets and our reconstruction error is more than 50% lower than the error from state-of-the-art models, which have to be trained for every mesh separately. Additionally, we visualize the underlying dynamics of unseen mesh sequences with an autoencoder trained on different classes of meshes.

Moritz Wolter, Felix Blanke, Raoul Heese et al.

As neural networks become able to generate realistic artificial images, they have the potential to improve movies, music, video games and make the internet an even more creative and inspiring place. Yet, the latest technology potentially enables new digital ways to lie. In response, the need for a diverse and reliable method toolbox arises to identify artificial images and other content. Previous work primarily relies on pixel-space CNNs or the Fourier transform. To the best of our knowledge, synthesized fake image analysis and detection methods based on a multi-scale wavelet representation, localized in both space and frequency, have been absent thus far. The wavelet transform conserves spatial information to a degree, which allows us to present a new analysis. Comparing the wavelet coefficients of real and fake images allows interpretation. Significant differences are identified. Additionally, this paper proposes to learn a model for the detection of synthetic images based on the wavelet-packet representation of natural and GAN-generated images. Our lightweight forensic classifiers exhibit competitive or improved performance at comparatively small network sizes, as we demonstrate on the FFHQ, CelebA and LSUN source identification problems. Furthermore, we study the binary FaceForensics++ fake-detection problem.

Yanying Zhou, Shijie Li, Jochen Garcke

When robots handle navigation tasks while avoiding collisions, they perform in crowded and complex environments not as good as in stable and homogeneous environments. This often results in a low success rate and poor efficiency. Therefore, we propose a novel Foresight Social-aware Reinforcement Learning (FSRL) framework for mobile robots to achieve collision-free navigation. Compared to previous learning-based methods, our approach is foresighted. It not only considers the current human-robot interaction to avoid an immediate collision, but also estimates upcoming social interactions to still keep distance in the future. Furthermore, an efficiency constraint is introduced in our approach that significantly reduces navigation time. Comparative experiments are performed to verify the effectiveness and efficiency of our proposed method under more realistic and challenging simulated environments.

Sara Hahner, Rodrigo Iza-Teran, Jochen Garcke

For sequences of complex 3D shapes in time we present a general approach to detect patterns for their analysis and to predict the deformation by making use of structural components of the complex shape. We incorporate long short-term memory (LSTM) layers into an autoencoder to create low dimensional representations that allow the detection of patterns in the data and additionally detect the temporal dynamics in the deformation behavior. This is achieved with two decoders, one for reconstruction and one for prediction of future time steps of the sequence. In a preprocessing step the components of the studied object are converted to oriented bounding boxes which capture the impact of plastic deformation and allow reducing the dimensionality of the data describing the structure. The architecture is tested on the results of 196 car crash simulations of a model with 133 different components, where material properties are varied. In the latent representation we can detect patterns in the plastic deformation for the different components. The predicted bounding boxes give an estimate of the final simulation result and their quality is improved in comparison to different baselines.

Skylar Sible, Rodrigo Iza-Teran, Jochen Garcke et al.

Modern product design in the engineering domain is increasingly driven by computational analysis including finite-element based simulation, computational optimization, and modern data analysis techniques such as machine learning. To apply these methods, suitable data representations for components under development as well as for related design criteria have to be found. While a component's geometry is typically represented by a polygon surface mesh, it is often not clear how to parametrize critical design properties in order to enable efficient computational analysis. In the present work, we propose a novel methodology to obtain a parameterization of a component's plastic deformation behavior under stress, which is an important design criterion in many application domains, for example, when optimizing the crash behavior in the automotive context. Existing parameterizations limit computational analysis to relatively simple deformations and typically require extensive input by an expert, making the design process time intensive and costly. Hence, we propose a way to derive a compact descriptor of deformation behavior that is based on spectral mesh processing and enables a low-dimensional representation of also complex deformations.We demonstrate the descriptor's ability to represent relevant deformation behavior by applying it in a nearest-neighbor search to identify similar simulation results in a filtering task. The proposed descriptor provides a novel approach to the parametrization of geometric deformation behavior and enables the use of state-of-the-art data analysis techniques such as machine learning to engineering tasks concerned with plastic deformation behavior.

Ribana Roscher, Bastian Bohn, Marco F. Duarte et al.

Machine learning methods have been remarkably successful for a wide range of application areas in the extraction of essential information from data. An exciting and relatively recent development is the uptake of machine learning in the natural sciences, where the major goal is to obtain novel scientific insights and discoveries from observational or simulated data. A prerequisite for obtaining a scientific outcome is domain knowledge, which is needed to gain explainability, but also to enhance scientific consistency. In this article we review explainable machine learning in view of applications in the natural sciences and discuss three core elements which we identified as relevant in this context: transparency, interpretability, and explainability. With respect to these core elements, we provide a survey of recent scientific works that incorporate machine learning and the way that explainable machine learning is used in combination with domain knowledge from the application areas.

Laura von Rueden, Sebastian Mayer, Katharina Beckh et al.

Despite its great success, machine learning can have its limits when dealing with insufficient training data. A potential solution is the additional integration of prior knowledge into the training process which leads to the notion of informed machine learning. In this paper, we present a structured overview of various approaches in this field. We provide a definition and propose a concept for informed machine learning which illustrates its building blocks and distinguishes it from conventional machine learning. We introduce a taxonomy that serves as a classification framework for informed machine learning approaches. It considers the source of knowledge, its representation, and its integration into the machine learning pipeline. Based on this taxonomy, we survey related research and describe how different knowledge representations such as algebraic equations, logic rules, or simulation results can be used in learning systems. This evaluation of numerous papers on the basis of our taxonomy uncovers key methods in the field of informed machine learning.

Friedrich Solowjow, Dominik Baumann, Jochen Garcke et al.

Common event-triggered state estimation (ETSE) algorithms save communication in networked control systems by predicting agents' behavior, and transmitting updates only when the predictions deviate significantly. The effectiveness in reducing communication thus heavily depends on the quality of the dynamics models used to predict the agents' states or measurements. Event-triggered learning is proposed herein as a novel concept to further reduce communication: whenever poor communication performance is detected, an identification experiment is triggered and an improved prediction model learned from data. Effective learning triggers are obtained by comparing the actual communication rate with the one that is expected based on the current model. By analyzing statistical properties of the inter-communication times and leveraging powerful convergence results, the proposed trigger is proven to limit learning experiments to the necessary instants. Numerical and physical experiments demonstrate that event-triggered learning improves robustness toward changing environments and yields lower communication rates than common ETSE.