Search

Scientific discoveries are often made by finding a pattern or object that was not predicted by the known rules of science. Oftentimes, these anomalous events or objects that do not conform to the norms are an indication that the rules of science governing the data are incomplete, and something new needs to be present to explain these unexpected outliers. The challenge of finding anomalies can be confounding since it requires codifying a complete knowledge of the known scientific behaviors and then projecting these known behaviors on the data to look for deviations. When utilizing machine learning, this presents a particular challenge since we require that the model not only understands scientific data perfectly but also recognizes when the data is inconsistent and out of the scope of its trained behavior. In this paper, we present three datasets aimed at developing machine learning-based anomaly detection for disparate scientific domains covering astrophysics, genomics, and polar science. We present the different datasets along with a scheme to make machine learning challenges around the three datasets findable, accessible, interoperable, and reusable (FAIR). Furthermore, we present an approach that generalizes to future machine learning challenges, enabling the possibility of large, more compute-intensive challenges that can ultimately lead to scientific discovery., Comment: 18 pages 6 figures to be submitted to Nature Communications

Time-parallel methods can reduce the wall clock time required for the accurate numerical solution of differential equations by parallelizing across the time-dimension. In this paper, we present and test the convergence behavior of a multiscale, micro-macro version of a Parareal method for stochastic differential equations (SDEs). In our method, the fine propagator of the SDE is based on a high-dimensional slow-fast microscopic model; the coarse propagator is based on a model-reduced version of the latter, that captures the low-dimensional, effective dynamics at the slow time scales. We investigate how the model error of the approximate model influences the convergence of the micro-macro Parareal algorithm and we support our analysis with numerical experiments. This is an extended and corrected version of [Domain Decomposition Methods in Science and Engineering XXVII. DD 2022, vol 149 (2024), pp. 69-76, Bossuyt, I., Vandewalle, S., Samaey, G.].

Large-scale 2D datasets have been instrumental in advancing machine learning; however, progress in 3D vision tasks has been relatively slow. This disparity is largely due to the limited availability of 3D benchmarking datasets. In particular, creating real-world point cloud datasets for indoor scene semantic segmentation presents considerable challenges, including data collection within confined spaces and the costly, often inaccurate process of per-point labeling to generate ground truths. While synthetic datasets address some of these challenges, they often fail to replicate real-world conditions, particularly the occlusions that occur in point clouds collected from real environments. Existing 3D benchmarking datasets typically evaluate deep learning models under the assumption that training and test data are independently and identically distributed (IID), which affects the models' usability for real-world point cloud segmentation. To address these challenges, we introduce the BelHouse3D dataset, a new synthetic point cloud dataset designed for 3D indoor scene semantic segmentation. This dataset is constructed using real-world references from 32 houses in Belgium, ensuring that the synthetic data closely aligns with real-world conditions. Additionally, we include a test set with data occlusion to simulate out-of-distribution (OOD) scenarios, reflecting the occlusions commonly encountered in real-world point clouds. We evaluate popular point-based semantic segmentation methods using our OOD setting and present a benchmark. We believe that BelHouse3D and its OOD setting will advance research in 3D point cloud semantic segmentation for indoor scenes, providing valuable insights for the development of more generalizable models., Comment: 20 pages, 6 figures, 3 tables, accepted at ECCV 2024 Workshops

We present an innovative robotic device designed to provide controlled motion for studying active matter. Motion is driven by an internal vibrator powered by a small rechargeable battery. The system integrates acoustic and magnetic sensors along with a programmable microcontroller. Unlike conventional vibrobots, the motor induces horizontal vibrations, resulting in cycloidal trajectories that have been characterized and optimized. Portions of these orbits can be utilized to create specific motion patterns. As a proof of concept, we demonstrate how this versatile system can be exploited to develop active particles with varying dynamics, ranging from ballistic motion to run-and-tumble diffusive behavior., Comment: 8 pages, 7 figures

Deep learning has shown tremendous progress in a wide range of digital pathology and medical image classification tasks. Its integration into safe clinical decision-making support requires robust and reliable models. However, real-world data comes with diversities that often lie outside the intended source distribution. Moreover, when test samples are dramatically different, clinical decision-making is greatly affected. Quantifying predictive uncertainty in models is crucial for well-calibrated predictions and determining when (or not) to trust a model. Unfortunately, many works have overlooked the importance of predictive uncertainty estimation. This paper evaluates whether predictive uncertainty estimation adds robustness to deep learning-based diagnostic decision-making systems. We investigate the effect of various carcinoma distribution shift scenarios on predictive performance and calibration. We first systematically investigate three popular methods for improving predictive uncertainty: Monte Carlo dropout, deep ensemble, and few-shot learning on lung adenocarcinoma classification as a primary disease in whole slide images. Secondly, we compare the effectiveness of the methods in terms of performance and calibration under clinically relevant distribution shifts such as in-distribution shifts comprising primary disease sub-types and other characterization analysis data; out-of-distribution shifts comprising well-differentiated cases, different organ origin, and imaging modality shifts. While studies on uncertainty estimation exist, to our best knowledge, no rigorous large-scale benchmark compares predictive uncertainty estimation including these dataset shifts for lung carcinoma classification., Comment: 17 pages, 2 figures, 5 tables

We propose a robust method for point cloud registration that can handle both unknown scales and extreme outlier ratios. Our method, dubbed PCR-99, uses a deterministic 3-point sampling approach with two novel mechanisms that significantly boost the speed: (1) an improved ordering of the samples based on pairwise scale consistency, prioritizing the point correspondences that are more likely to be inliers, and (2) an efficient outlier rejection scheme based on triplet scale consistency, prescreening bad samples and reducing the number of hypotheses to be tested. Our evaluation shows that, up to 98% outlier ratio, the proposed method achieves comparable performance to the state of the art. At 99% outlier ratio, however, it outperforms the state of the art for both known-scale and unknown-scale problems. Especially for the latter, we observe a clear superiority in terms of robustness and speed., Comment: Accepted to ECCV 2024 Workshop on Recovering 6D Object Pose (R6D)

Event of the same type occurring several times for one individual (recurrent events) are present in various domains (industrial systems reliability, episodes of unemployment, political conflicts, chronic diseases episodes). Analysis of such kind of data should account for the whole recurrence process dynamics rather than only focusing on the number of observed events. Statistical models for recurrent events analysis are developed in the counting process probabilistic framework. One of the often-used models is the Andersen-Gill model, a generalization of the well-known Cox model for durations, which assumes that the baseline intensity of the recurrence process is time-dependent and is adjusted for covariates. For an individual i with covariates Xi, the intensity is as follows: $\lambda$_{ik}(t;$\theta$) = $\lambda$_0(t) exp (X_i $\beta$). The baseline intensity can be specified parametrically, in a form of Weibull: $\lambda$_0 (t) = $\gamma$_{1} $\gamma$_{2} t^{$\gamma$_2-1}, with $\gamma$1 scale parameter et $\gamma$2 shape parameter. However, the observed covariates are often insufficient to explain the observed heterogeneity in data. This is often the case of clinical trials data containing information on patients. In this article a mixture model for recurrent events analysis is proposed. This model allows to account for unobserved heterogeneity and to cluster individuals according to their recurrence process. The intensity of the process is parametrically specified within each class and depend on observed covariates. Thus, the intensity becomes specific to class k: $\lambda$_{ik} (t; $\theta$_k) = $\gamma$_{1k} $\gamma$_{2k} t^{$\gamma$_{2k}-1} exp (X_i $\beta$_k). The model parameters are estimated by the Maximum Likelihood method, using the EM algorithm. The BIC criterion is employed to choose the optimal number of classes. Model feasibility is verified by Monte Carlo simulations. An application to real data concerning hospital readmissions of elderly patients is proposed. The proposed model feasibility is empirically verified (the optimization algorithm converges, providing non-biased estimates). The real data application allows to identify two clinically relevant classes of patients., Comment: in French, 54es Journ{\'e}es de Statistique, Jul 2023, Bruxelles (BEL), France

In this paper, we are concerned with the micro-macro Parareal algorithm for the simulation of initial-value problems. In this algorithm, a coarse (fast) solver is applied sequentially over the time domain, and a fine (time-consuming) solver is applied as a corrector in parallel over smaller chunks of the time interval. Moreover, the coarse solver acts on a reduced state variable, which is coupled to the fine state variable through appropriate coupling operators. We first provide a contribution to the convergence analysis of the micro-macro Parareal method for multiscale linear ordinary differential equations (ODEs). Then, we extend a variant of the micro-macro Parareal algorithm for scalar stochastic differential equations (SDEs) to higher-dimensional SDEs.

In nature, metachronal coordination is an efficient strategy for fluid pumping and self-propulsion. Yet, mimetic systems for this type of organization are scarce. Recently, metachronal motion was observed in a bead-based magnetocapillary mesoswimmer, but the mechanism of such device's behavior remained unclear. Here, we combine theory and experiments to explain two swimming regimes that we identify by modulation of the driving frequency. In the linear, low-frequency regime, the swimmer motion originates from individual bead rotations. However, the high-frequency, metachronal regime is dominated by deformations of the device near a mechanical resonance, which highlights the role of bead inertia for optimized self-propulsion., Comment: Article: 6 pages, 3 figures; Supplementary information: 5 pages, 8 figures

We propose a micro-macro parallel-in-time Parareal method for scalar McKean-Vlasov stochastic differential equations (SDEs). In the algorithm, the fine Parareal propagator is a Monte Carlo simulation of an ensemble of particles, while an approximate ordinary differential equation (ODE) description of the mean and the variance of the particle distribution is used as a coarse Parareal propagator to achieve speedup. We analyse the convergence behaviour of our method for a linear problem and provide numerical experiments indicating the parallel weak scaling of the algorithm on a set of examples. We show, with numerical experiments, that convergence typically takes place in a low number of iterations, depending on the quality of the ODE predictor. For bimodal SDEs, we avoid quality deterioration of the coarse predictor (compared to unimodal SDEs) through the usage of multiple ODEs, each describing the mean and variance of the particle distribution in locally unimodal regions of the phase space. The benefit of the proposed algorithm can be viewed through two lenses: (i) through the parallel-in-time lens, speedup is obtained through the use of a very cheap coarse integrator (an ODE moment model), and (ii) through the moment models lens, accuracy is iteratively gained through the use of parallel machinery as a corrector. In contrast to the isolated use of a moment model, the proposed method (iteratively) converges to the true distribution generated by the SDE.

In the framework of model-based clustering, a model, called multi-partitions clustering, allowing several latent class variables has been proposed. This model assumes that the distribution of the observed data can be factorized into several independent blocks of variables, each block following its own mixture model. In this paper, we assume that each block follows a non parametric latent class model, {\it i.e.} independence of the variables in each component of the mixture with no parametric assumption on their class conditional distribution. The purpose is to deduce, from the observation of a sample, the number of blocks, the partition of the variables into the blocks and the number of components in each block, which characterise the proposed model. By following recent literature on model and variable selection in non-parametric mixture models, we propose to discretize the data into bins. This permits to apply the classical multi-partition clustering procedure for parametric multinomials, which are based on a penalized likelihood method (\emph{e.g.} BIC). The consistency of the procedure is obtained and an efficient optimization is proposed. The performances of the model are investigated on simulated data.

Recently, transformer networks have outperformed traditional deep neural networks in natural language processing and show a large potential in many computer vision tasks compared to convolutional backbones. In the original transformer, readout tokens are used as designated vectors for aggregating information from other tokens. However, the performance of using readout tokens in a vision transformer is limited. Therefore, we propose a novel fusion strategy to integrate radar data into a dense prediction transformer network by reassembling camera representations with radar representations. Instead of using readout tokens, radar representations contribute additional depth information to a monocular depth estimation model and improve performance. We further investigate different fusion approaches that are commonly used for integrating additional modality in a dense prediction transformer network. The experiments are conducted on the nuScenes dataset, which includes camera images, lidar, and radar data. The results show that our proposed method yields better performance than the commonly used fusion strategies and outperforms existing convolutional depth estimation models that fuse camera images and radar., Comment: 5 pages, 2 figures and 1 table, accepted to ICASSP2023