Search

Traditionally, 3d indoor datasets have generally prioritized scale over ground-truth accuracy in order to obtain improved generalization. However, using these datasets to evaluate dense geometry tasks, such as depth rendering, can be problematic as the meshes of the dataset are often incomplete and may produce wrong ground truth to evaluate the details. In this paper, we propose SCRREAM, a dataset annotation framework that allows annotation of fully dense meshes of objects in the scene and registers camera poses on the real image sequence, which can produce accurate ground truth for both sparse 3D as well as dense 3D tasks. We show the details of the dataset annotation pipeline and showcase four possible variants of datasets that can be obtained from our framework with example scenes, such as indoor reconstruction and SLAM, scene editing & object removal, human reconstruction and 6d pose estimation. Recent pipelines for indoor reconstruction and SLAM serve as new benchmarks. In contrast to previous indoor dataset, our design allows to evaluate dense geometry tasks on eleven sample scenes against accurately rendered ground truth depth maps.

The present study investigates the behaviour of the far-field sound radiated by low Mach number tip clearance flow induced by placing a stationary cambered airfoil adjacent to a stationary wall. The tip clearance heights ranged from 14% to 30% of the incoming, undisturbed boundary layer thickness and the clearance heights based Reynolds numbers were between 2,600 and 16,000. The far-field sound measured using a microphone array was beamformed to reveal the dominant noise sources and how they behave when the flow Mach number, angle of attack and the clearance height were varied. The near-field behaviour was also examined through PIV measurements and surface pressure fluctuation measurements on the tip. The results show that the mid-to-high frequency noise generated by tip clearances is dominated by the leakage flow in the mid-chord and leading-edge regions, while a distinct low-frequency noise source with a different scaling behaviour exists close to the trailing-edge of the tip clearance. The origin of this low-frequency noise source is believed to be the tip separation vortex that resides close to the trailing-edge and induces significant turbulence levels in the region. The strength of this noise source decreases with clearance height which is consistent with a reduction in turbulence levels associated with the separation vortex. The magnitude of the mid-frequency clearance noise which scales with the sixth power of the Mach number, decreases with tip clearance height due to a reduction in the fluctuating pressure on the airfoil tip surface. The time-scale of this sound was independent of the flow velocity, implying that the source is non-compact. Smaller tip clearances were also found to generate louder high-frequency noise due to intense turbulence and pressure fluctuation levels concentrated near the leading-edge of the clearance., Comment: Submitted to Applied Acoustics Journal

This paper introduces AMMORE, a new dataset of 53,000 math open-response question-answer pairs from Rori, a learning platform used by students in several African countries and conducts two experiments to evaluate the use of large language models (LLM) for grading particularly challenging student answers. The AMMORE dataset enables various potential analyses and provides an important resource for researching student math acquisition in understudied, real-world, educational contexts. In experiment 1 we use a variety of LLM-driven approaches, including zero-shot, few-shot, and chain-of-thought prompting, to grade the 1% of student answers that a rule-based classifier fails to grade accurately. We find that the best-performing approach -- chain-of-thought prompting -- accurately scored 92% of these edge cases, effectively boosting the overall accuracy of the grading from 98.7% to 99.9%. In experiment 2, we aim to better understand the consequential validity of the improved grading accuracy, by passing grades generated by the best-performing LLM-based approach to a Bayesian Knowledge Tracing (BKT) model, which estimated student mastery of specific lessons. We find that relatively modest improvements in model accuracy at the individual question level can lead to significant changes in the estimation of student mastery. Where the rules-based classifier currently used to grade student, answers misclassified the mastery status of 6.9% of students across their completed lessons, using the LLM chain-of-thought approach this misclassification rate was reduced to 2.6% of students. Taken together, these findings suggest that LLMs could be a valuable tool for grading open-response questions in K-12 mathematics education, potentially enabling encouraging wider adoption of open-ended questions in formative assessment.

Nighttime camera-based depth estimation is a highly challenging task, especially for autonomous driving applications, where accurate depth perception is essential for ensuring safe navigation. We aim to improve the reliability of perception systems at night time, where models trained on daytime data often fail in the absence of precise but costly LiDAR sensors. In this work, we introduce Light Enhanced Depth (LED), a novel cost-effective approach that significantly improves depth estimation in low-light environments by harnessing a pattern projected by high definition headlights available in modern vehicles. LED leads to significant performance boosts across multiple depth-estimation architectures (encoder-decoder, Adabins, DepthFormer) both on synthetic and real datasets. Furthermore, increased performances beyond illuminated areas reveal a holistic enhancement in scene understanding. Finally, we release the Nighttime Synthetic Drive Dataset, a new synthetic and photo-realistic nighttime dataset, which comprises 49,990 comprehensively annotated images., Comment: Preprint. Code and dataset available on the project page : https://simondemoreau.github.io/LED/

The limited availability of accurate experimental data in wide ranges of pressure, temperature, and composition is the main constraining factor for the proper development and assessment of thermodynamic models and equations of state. In the particular case of carbon capture and storage (CCS) processes, there is a clear need for data sets related to the (carbon dioxide + oxygen) mixtures that this work aims to address. This work provides new experimental ($p$, $\rho$, $T$) data for three binary ($CO_{2}$ + $O_{2}$) mixtures with mole fractions of oxygen $x$($O_{2}$) = (0.05, 0.10, and 0.20) mol/mol, in the temperature range $T$ = (250 to 375) K and pressure range $p$ = (0.5 to 13) MPa. The measurements were performed with a high-precision single-sinker densimeter with magnetic suspension coupling. The density data were obtained with estimated expanded relative uncertainties of 0.02% for the highest densities and up to 0.3% for the lowest ones.The results were compared to the corresponding results calculated by the current reference equations of state for this kind of mixtures, namely the EOS-CG (combustion gases) and the GERG-2008 equation of state, respectively. The EOS-CG yields better estimations in density than the GERG-2008 equation of state. The results from the EOS-GC model show no systematic temperature dependence. For the GERG-2008 model, however, this criterion is significantly less fulfilled., Comment: arXiv admin note: substantial text overlap with arXiv:2409.06312

This work aims to address the technical aspects related to the thermodynamic characterization of natural gas mixtures blended with hydrogen for the introduction of alternative energy sources within the Power-to-Gas framework. For that purpose, new experimental speed of sound data are presented in the pressure range between (0.1 up to 13) MPa and at temperatures of (260, 273.16, 300, 325, and 350) K for two mixtures qualified as primary calibration standards: a 11 component synthetic natural gas mixture (11 M), and another low-calorific $H_{2}$-enriched natural gas mixture with a nominal molar percentage $x_{H_{2}}$ = 3%. Measurements have been gathered using a spherical acoustic resonator with an experimental expanded ($k$ = 2) uncertainty better than 200 parts in $10^{6}$ (0.02%) in the speed of sound. The heat capacity ratio as perfect-gas $\gamma_{pg}$, the molar heat capacity as perfect-gas $C_{p,m}^{pg}$, and the second $\beta_{a}$ and third $\gamma_{a}$ acoustic virial coefficients are derived from the speed of sound values. All the results are compared with the reference mixture models for natural gas-like mixtures, the AGA8-DC92 EoS and the GERG-2008 EoS, with special attention to the impact of hydrogen on those properties. Data are found to be mostly consistent within the model uncertainty in the 11 M synthetic mixture as expected, but for the hydrogen-enriched mixture in the limit of the model uncertainty at the highest measuring pressures.

This work aims to address the technical concerns related to the thermodynamic characterization of gas mixtures blended with hydrogen for the implementation of hydrogen as a new energy vector. For this purpose, new experimental speed of sound measurements have been done in gaseous and supercritical phases of two binary mixtures of nitrogen and hydrogen using the most accurate technique available, i.e., the spherical acoustic resonator, yielding an experimental expanded ($k$ = 2) uncertainty of only 220 parts in $10^{6}$ (0.022%). The measurements cover the pressure range between (0.5 and 20) MPa, the temperature range between (260 and 350) K, and the composition range with a nominal mole percentage of hydrogen of (5 and 10) mol%, respectively. From the speed of sound data sets, thermophysical properties that are relevant for the characterization of the mixture, namely the second $\beta_{a}$ and third $\gamma_{a}$ acoustic virial coefficients, are derived. These results are thoroughly compared and discussed with the established reference mixture models valid for mixtures of nitrogen and hydrogen, such as the AGA8-DC92 EoS, the GERG-2008 EoS, and the recently developed adaptation of the GERG-2008 EoS, here denoted GERG-$H_{2}$_improved EoS. Special attention has been given to the effect of hydrogen concentration on those properties, showing that only the GERG-$H_{2}$_improved EoS is consistent with the data sets within the experimental uncertainty in most measuring conditions.

The accurate knowledge of the thermophysical and thermodynamic properties of pure hydrogen and hydrogen mixtures plays an important role in the design and operation of many processes involved in hydrogen production, transport, storage, and use. These data are needed for the development of theoretical models necessary for the introduction of hydrogen as a promising energy carrier in the near future. A literature survey on both the available experimental data and the theoretical models associated with the thermodynamic properties of hydrogen mixtures, within the operational ranges of industrial interest for composition, temperature, and pressure, is presented in this work. Considering the available experimental data and the requirements for the design and operation of hydrogen systems, the most relevant gaps in temperature, pressure and composition are identified.

Modern machine learning relies nearly exclusively on dedicated electronic hardware accelerators. Photonic approaches, with low consumption and high operation speed, are increasingly considered for inference but, to date, remain mostly limited to relatively basic tasks. Simultaneously, the problem of training deep and complex neural networks, overwhelmingly performed through backpropagation, remains a significant limitation to the size and, consequently, the performance of current architectures and a major compute and energy bottleneck. Here, we experimentally implement a versatile and scalable training algorithm, called direct feedback alignment, on a hybrid electronic-photonic platform. An optical processing unit performs large-scale random matrix multiplications, which is the central operation of this algorithm, at speeds up to 1500 TeraOps. We perform optical training of one of the most recent deep learning architectures, including Transformers, with more than 1B parameters, and obtain good performances on both language and vision tasks. We study the compute scaling of our hybrid optical approach, and demonstrate a potential advantage for ultra-deep and wide neural networks, thus opening a promising route to sustain the exponential growth of modern artificial intelligence beyond traditional von Neumann approaches., Comment: 19 pages, 4 figures

JINet is a web browser-based platform intended to democratise access to advanced clinical and genomic data analysis software. It hosts numerous data analysis applications that are run in the safety of each User's web browser, without the data ever leaving their machine. JINet promotes collaboration, standardisation and reproducibility by sharing scripts rather than data and creating a self-sustaining community around it in which Users and data analysis tools developers interact thanks to JINets interoperability primitives., Comment: 13 pages, 6 figures, 1 table

Porous media are ubiquitous in the fields of energy storage and conversion, catalysis, biomechanics, hydrogeology, and other chemical engineering processes. These media possess high surface-to-volume ratios and their complex channels can restrict and direct the flow. This makes them key components of multiple equipment despite the challenges in selecting design parameters for specific applications. Pore-resolved CFD reveals the effects of their structure at the microscopic scale, but is currently limited by high computing costs and the performance of mesh generation algorithms. This work introduces a RBF-based representation of solids in a massively parallel immersed-boundary framework, enabling both the usage of non-conformal grids and dynamic mesh adaptation. We verify it using the method of manufactured solutions. We validate it using pressure drop measurements through porous silicone monoliths digitized by X-ray computed microtomography for Reynolds numbers up to 30, using grids of 200 M cells distributed over 8 k cores. The reliable model is then used to highlight that pore network structure is the main factor describing pressure evolution and that preferential channels are present at this scale of the porous media. This work opens the door to design and optimize processes by linking microscopic flow to macroscopic properties through the usage of physics-based digital twins of complex porous media., Comment: 26 pages, 18 figures

In the drug discovery process, where experiments can be costly and time-consuming, computational models that predict drug-target interactions are valuable tools to accelerate the development of new therapeutic agents. Estimating the uncertainty inherent in these neural network predictions provides valuable information that facilitates optimal decision-making when risk assessment is crucial. However, such models can be poorly calibrated, which results in unreliable uncertainty estimates that do not reflect the true predictive uncertainty. In this study, we compare different metrics, including accuracy and calibration scores, used for model hyperparameter tuning to investigate which model selection strategy achieves well-calibrated models. Furthermore, we propose to use a computationally efficient Bayesian uncertainty estimation method named Bayesian Linear Probing (BLP), which generates Hamiltonian Monte Carlo (HMC) trajectories to obtain samples for the parameters of a Bayesian Logistic Regression fitted to the hidden layer of the baseline neural network. We report that BLP improves model calibration and achieves the performance of common uncertainty quantification methods by combining the benefits of uncertainty estimation and probability calibration methods. Finally, we show that combining post hoc calibration method with well-performing uncertainty quantification approaches can boost model accuracy and calibration.

Unsupervised Domain Adaptation (DA) consists of adapting a model trained on a labeled source domain to perform well on an unlabeled target domain with some data distribution shift. While many methods have been proposed in the literature, fair and realistic evaluation remains an open question, particularly due to methodological difficulties in selecting hyperparameters in the unsupervised setting. With SKADA-Bench, we propose a framework to evaluate DA methods and present a fair evaluation of existing shallow algorithms, including reweighting, mapping, and subspace alignment. Realistic hyperparameter selection is performed with nested cross-validation and various unsupervised model selection scores, on both simulated datasets with controlled shifts and real-world datasets across diverse modalities, such as images, text, biomedical, and tabular data with specific feature extraction. Our benchmark highlights the importance of realistic validation and provides practical guidance for real-life applications, with key insights into the choice and impact of model selection approaches. SKADA-Bench is open-source, reproducible, and can be easily extended with novel DA methods, datasets, and model selection criteria without requiring re-evaluating competitors. SKADA-Bench is available on GitHub at https://github.com/scikit-adaptation/skada-bench.

This study develops a robust movie recommendation system using various machine learning techniques, including Non- Negative Matrix Factorization (NMF), Truncated Singular Value Decomposition (SVD), and K-Means clustering. The primary objective is to enhance user experience by providing personalized movie recommendations. The research encompasses data preprocessing, model training, and evaluation, highlighting the efficacy of the employed methods. Results indicate that the proposed system achieves high accuracy and relevance in recommendations, making significant contributions to the field of recommendations systems., Comment: Accepted by 2024 4th International Symposium on Computer Technology and Information Science, IEEE

We study the positive logic FO+ on finite words, and its fragments, pursuing and refining the work initiated in [Kuperberg 2023]. First, we transpose notorious logic equivalences into positive first-order logic: FO+ is equivalent to LTL+ , and its two-variable fragment FO2+ with (resp. without) successor available is equivalent to UTL+ with (resp. without) the "next" operator X available. This shows that despite previous negative results, the class of FO+-definable languages exhibits some form of robustness. We then exhibit an example of an FO-definable monotone language on one predicate, that is not FO+-definable, refining the example from [Kuperberg 2023] with 3 predicates. Moreover, we show that such a counter-example cannot be FO2-definable.

Physiological signal analysis often involves identifying events crucial to understanding biological dynamics. Traditional methods rely on handcrafted procedures or supervised learning, presenting challenges such as expert dependence, lack of robustness, and the need for extensive labeled data. Data-driven methods like Convolutional Dictionary Learning (CDL) offer an alternative but tend to produce spurious detections. This work introduces UNHaP (Unmix Noise from Hawkes Processes), a novel approach addressing the joint learning of temporal structures in events and the removal of spurious detections. Leveraging marked Hawkes processes, UNHaP distinguishes between events of interest and spurious ones. By treating the event detection output as a mixture of structured and unstructured events, UNHaP efficiently unmixes these processes and estimates their parameters. This approach significantly enhances the understanding of event distributions while minimizing false detection rates.

In this work, we introduce a quantum-based quantitative phase microscopy technique using a phase gradient approach that is inherently background resistant and does not rely on interferometry or scanning. Here, a transparent sample is illuminated by both photons of a position-momentum entangled pair with one photon setup for position measurement in the near-field (NF) of the sample and its partner for momentum measurement in the far-field (FF). By virtue of the spatial correlation property inherent to the entanglement, both the position and momentum information of the photons can thus be obtained simultaneously. The phase profile of the sample is then deduced through a phase gradient measurement obtained by measuring the centroid shift of the photons' in the FF momentum plane for each NF position. We show that the technique, while achieving an imaging resolution of 2.76\,$\mu$m, is phase accurate to at least $\lambda/30$ and phase sensitive to $\lambda/100$ at a wavelength of 810\,nm. In addition, through the temporal correlation between the photon pairs, our technique shows resilience to strong dynamic background lights, which can prove difficult to account for in classical phase imaging techniques. We believe this work marks a significant advancement in the capabilities of quantum phase microscopy and quantum imaging in general, it showcases imaging and phase resolutions approaching those attainable with classical phase microscopes. This advancement brings quantum imaging closer to practical real-world applications, heralding new possibilities in the field.

Many modern spatio-temporal data sets, in sociology, epidemiology or seismology, for example, exhibit self-exciting characteristics, triggering and clustering behaviors both at the same time, that a suitable Hawkes space-time process can accurately capture. This paper aims to develop a fast and flexible parametric inference technique to recover the parameters of the kernel functions involved in the intensity function of a space-time Hawkes process based on such data. Our statistical approach combines three key ingredients: 1) kernels with finite support are considered, 2) the space-time domain is appropriately discretized, and 3) (approximate) precomputations are used. The inference technique we propose then consists of a $\ell_2$ gradient-based solver that is fast and statistically accurate. In addition to describing the algorithmic aspects, numerical experiments have been carried out on synthetic and real spatio-temporal data, providing solid empirical evidence of the relevance of the proposed methodology.

PLATO (PLAnetary Transits and Oscillations of stars) is ESA's M3 mission designed to detect and characterise extrasolar planets and perform asteroseismic monitoring of a large number of stars. PLATO will detect small planets (down to <2 R_(Earth)) around bright stars (<11 mag), including terrestrial planets in the habitable zone of solar-like stars. With the complement of radial velocity observations from the ground, planets will be characterised for their radius, mass, and age with high accuracy (5 %, 10 %, 10 % for an Earth-Sun combination respectively). PLATO will provide us with a large-scale catalogue of well-characterised small planets up to intermediate orbital periods, relevant for a meaningful comparison to planet formation theories and to better understand planet evolution. It will make possible comparative exoplanetology to place our Solar System planets in a broader context. In parallel, PLATO will study (host) stars using asteroseismology, allowing us to determine the stellar properties with high accuracy, substantially enhancing our knowledge of stellar structure and evolution. The payload instrument consists of 26 cameras with 12cm aperture each. For at least four years, the mission will perform high-precision photometric measurements. Here we review the science objectives, present PLATO's target samples and fields, provide an overview of expected core science performance as well as a description of the instrument and the mission profile at the beginning of the serial production of the flight cameras. PLATO is scheduled for a launch date end 2026. This overview therefore provides a summary of the mission to the community in preparation of the upcoming operational phases.

We report a theoretical and experimental investigation of fiber Fabry-Perot cavities aimed at enhancing Kerr frequency comb generation. The modulation instability (MI) power threshold is derived from the linear stability analysis of a generalized Lugiato-Lefever equation. By combining this analysis with the concepts of power enhancement factor (PEF) and optimal coupling, we predict the ideal manufacturing parameters of fiber Fabry-Perot (FFP) cavities for the MI Kerr frequency comb generation. Our findings reveal a distinction between the optimal coupling for modulation instability and that of the cold cavity. Consequently, mirror reflectivity must be adjusted to suit the specific application. We verified the predictions of our theory by measuring the MI power threshold as a function of detuning for three different cavities.

MISTRAL is the new Faint Object Spectroscopic Camera mounted at the folded Cassegrain focus of the 1.93m telescope of Haute-Provence Observatory. We describe the design and components of the instrument and give some details about its operation. We emphasise in particular the various observing modes and the performances of the detector. A short description is also given about the working environment. Various types of objects, including stars, nebulae, comets, novae, galaxies have been observed during various test phases to evaluate the performances of the instrument. The instrument covers the range of 4000 to 8000A with the blue setting, or from 6000 to 10000A with the red setting, at an average spectral resolution of 700. Its peak efficiency is about 22% at 6000A. In spectroscopy, a limiting magnitude of 19.5 can be achieved for a point source in one hour with a signal to noise of 3 in the continuum (and better if emission lines are present). In imaging mode, limiting magnitudes of 20-21 can be obtained in 10-20mn (with average seing conditions of 2.5 arcsec at OHP). The instrument is very users-friendly and can be put into operations in less than 15mn (rapid change-over from the other instrument in use) if required by the science (like for Gamma-Rays Bursts). Some first scientific results are described for various types of objects, and in particular for the follow-up of GRBs. While some further improvements are still under way, in particular to ease the switch from blue to red setting and add more grisms or filters, MISTRAL is ready for the follow-up of transients and other variable objects, in the soon-to-come era of e.g. the SVOM satellite and of the Rubin telescope., Comment: Accepted in A&A

Despite the increasing interest in enhancing perception systems for autonomous vehicles, the online calibration between event cameras and LiDAR - two sensors pivotal in capturing comprehensive environmental information - remains unexplored. We introduce MULi-Ev, the first online, deep learning-based framework tailored for the extrinsic calibration of event cameras with LiDAR. This advancement is instrumental for the seamless integration of LiDAR and event cameras, enabling dynamic, real-time calibration adjustments that are essential for maintaining optimal sensor alignment amidst varying operational conditions. Rigorously evaluated against the real-world scenarios presented in the DSEC dataset, MULi-Ev not only achieves substantial improvements in calibration accuracy but also sets a new standard for integrating LiDAR with event cameras in mobile platforms. Our findings reveal the potential of MULi-Ev to bolster the safety, reliability, and overall performance of event-based perception systems in autonomous driving, marking a significant step forward in their real-world deployment and effectiveness., Comment: Accepted at CVPR 2024 Workshop on Autonomous Driving. Copyright 2024 IEEE

Dissipative environments are ubiquitous in nature, from microscopic swimmers in low-Reynolds-number fluids to macroscopic animals in frictional media. In this study, motivated by various behaviours of {\it Caenorhabditis elegans} during swimming and crawling locomotion, we consider a mathematical model of a slender elastic locomotor with an internal rhythmic neural pattern generator. By analysing the dynamical systems of the model using a Poincar\'e section, we found that local neuromechanical adjustments to the environment can create robust undulatory locomotion. This progressive behaviour emerges as a global stable periodic orbit in a broad range of parameter regions. Further, by controlling the mechanosensation, we were able to design the dynamical systems to manoeuvre with progressive, reverse, and turning motions as well as apparently random, complex behaviours, as experimentally observed in {\it C. elegans}. The mechanisms found in this study, together with our methodologies with the dynamical systems viewpoint, are useful for deciphering complex animal adaptive behaviours and also designing adaptive robots for a wide range of dissipative environments., Comment: 13 pages, 9 figures

Objective. This study conduct an extensive Brain-computer interfaces (BCI) reproducibility analysis on open electroencephalography datasets, aiming to assess existing solutions and establish open and reproducible benchmarks for effective comparison within the field. The need for such benchmark lies in the rapid industrial progress that has given rise to undisclosed proprietary solutions. Furthermore, the scientific literature is dense, often featuring challenging-to-reproduce evaluations, making comparisons between existing approaches arduous. Approach. Within an open framework, 30 machine learning pipelines (separated into raw signal: 11, Riemannian: 13, deep learning: 6) are meticulously re-implemented and evaluated across 36 publicly available datasets, including motor imagery (14), P300 (15), and SSVEP (7). The analysis incorporates statistical meta-analysis techniques for results assessment, encompassing execution time and environmental impact considerations. Main results. The study yields principled and robust results applicable to various BCI paradigms, emphasizing motor imagery, P300, and SSVEP. Notably, Riemannian approaches utilizing spatial covariance matrices exhibit superior performance, underscoring the necessity for significant data volumes to achieve competitive outcomes with deep learning techniques. The comprehensive results are openly accessible, paving the way for future research to further enhance reproducibility in the BCI domain. Significance. The significance of this study lies in its contribution to establishing a rigorous and transparent benchmark for BCI research, offering insights into optimal methodologies and highlighting the importance of reproducibility in driving advancements within the field., Comment: 43 pages, 13 figures, 5 tables

Identifying the chemical structure from a graphical representation, or image, of a molecule is a challenging pattern recognition task that would greatly benefit drug development. Yet, existing methods for chemical structure recognition do not typically generalize well, and show diminished effectiveness when confronted with domains where data is sparse, or costly to generate, such as hand-drawn molecule images. To address this limitation, we propose a new chemical structure recognition tool that delivers state-of-the-art performance and can adapt to new domains with a limited number of data samples and supervision. Unlike previous approaches, our method provides atom-level localization, and can therefore segment the image into the different atoms and bonds. Our model is the first model to perform OCSR with atom-level entity detection with only SMILES supervision. Through rigorous and extensive benchmarking, we demonstrate the preeminence of our chemical structure recognition approach in terms of data efficiency, accuracy, and atom-level entity prediction., Comment: Accepted in IEEE/CVF Conference on Computer Vision and Pattern Recognition 2024

Diffusion-weighted MRI (DWI) is essential for stroke diagnosis, treatment decisions, and prognosis. However, image and disease variability hinder the development of generalizable AI algorithms with clinical value. We address this gap by presenting a novel ensemble algorithm derived from the 2022 Ischemic Stroke Lesion Segmentation (ISLES) challenge. ISLES'22 provided 400 patient scans with ischemic stroke from various medical centers, facilitating the development of a wide range of cutting-edge segmentation algorithms by the research community. Through collaboration with leading teams, we combined top-performing algorithms into an ensemble model that overcomes the limitations of individual solutions. Our ensemble model achieved superior ischemic lesion detection and segmentation accuracy on our internal test set compared to individual algorithms. This accuracy generalized well across diverse image and disease variables. Furthermore, the model excelled in extracting clinical biomarkers. Notably, in a Turing-like test, neuroradiologists consistently preferred the algorithm's segmentations over manual expert efforts, highlighting increased comprehensiveness and precision. Validation using a real-world external dataset (N=1686) confirmed the model's generalizability. The algorithm's outputs also demonstrated strong correlations with clinical scores (admission NIHSS and 90-day mRS) on par with or exceeding expert-derived results, underlining its clinical relevance. This study offers two key findings. First, we present an ensemble algorithm (https://github.com/Tabrisrei/ISLES22_Ensemble) that detects and segments ischemic stroke lesions on DWI across diverse scenarios on par with expert (neuro)radiologists. Second, we show the potential for biomedical challenge outputs to extend beyond the challenge's initial objectives, demonstrating their real-world clinical applicability.

The interaction of a gravitational wave (GW) with an elastic body is usually described in terms of a GW "force" driving the oscillations of the body's normal modes. However, this description is only possible for GW frequencies for which the response of the elastic body is dominated by a few normal modes. At higher frequencies the normal modes blend into a quasi-continuum and a field-theoretical description, as pioneered by Dyson already in 1969, becomes necessary. However, since the metric perturbation $h_{\mu\nu}$ is an intrinsically relativistic object, a consistent coupling to GWs can only be obtained within a relativistic (and, in fact generally covariant) theory of elasticity. We develop such a formalism using the methods of modern effective field theories, and we use it to provide a derivation of the interaction of elastic bodies with GWs valid also in the high-frequency regime, providing a first-principle derivation of Dyson's result (and partially correcting it). We also stress that the field-theoretical results are obtained working in the TT frame, while the description in terms of a force driving the normal modes is only valid in the proper detector frame. We show how to transform the results between the two frames. Beside an intrinsic conceptual interest, these results are relevant to the computation of the sensitivity of the recently proposed Lunar Gravitational Wave Antenna., Comment: 63 pages

In modern medical diagnostics, magnetic resonance imaging (MRI) is an important technique that provides detailed insights into anatomical structures. In this paper, we present a comprehensive methodology focusing on streamlining the segmentation, reconstruction, and visualization process of 3D MRI data. Segmentation involves the extraction of anatomical regions with the help of state-of-the-art deep learning algorithms. Then, 3D reconstruction converts segmented data from the previous step into multiple 3D representations. Finally, the visualization stage provides efficient and interactive presentations of both 2D and 3D MRI data. Integrating these three steps, the proposed system is able to augment the interpretability of the anatomical information from MRI scans according to our interviews with doctors. Even though this system was originally designed and implemented as part of human brain haptic feedback simulation for surgeon training, it can also provide experienced medical practitioners with an effective tool for clinical data analysis, surgical planning and other purposes

Motivated by the challenge of seamless cross-dataset transfer in EEG signal processing, this article presents an exploratory study on the use of Joint Embedding Predictive Architectures (JEPAs). In recent years, self-supervised learning has emerged as a promising approach for transfer learning in various domains. However, its application to EEG signals remains largely unexplored. In this article, we introduce Signal-JEPA for representing EEG recordings which includes a novel domain-specific spatial block masking strategy and three novel architectures for downstream classification. The study is conducted on a 54 subjects dataset and the downstream performance of the models is evaluated on three different BCI paradigms: motor imagery, ERP and SSVEP. Our study provides preliminary evidence for the potential of JEPAs in EEG signal encoding. Notably, our results highlight the importance of spatial filtering for accurate downstream classification and reveal an influence of the length of the pre-training examples but not of the mask size on the downstream performance.

Reliable multimodal sensor fusion algorithms require accurate spatiotemporal calibration. Recently, targetless calibration techniques based on implicit neural representations have proven to provide precise and robust results. Nevertheless, such methods are inherently slow to train given the high computational overhead caused by the large number of sampled points required for volume rendering. With the recent introduction of 3D Gaussian Splatting as a faster alternative to implicit representation methods, we propose to leverage this new rendering approach to achieve faster multi-sensor calibration. We introduce 3DGS-Calib, a new calibration method that relies on the speed and rendering accuracy of 3D Gaussian Splatting to achieve multimodal spatiotemporal calibration that is accurate, robust, and with a substantial speed-up compared to methods relying on implicit neural representations. We demonstrate the superiority of our proposal with experimental results on sequences from KITTI-360, a widely used driving dataset., Comment: Accepted at IROS 2024 (Oral presentation). Project page: https://qherau.github.io/3DGS-Calib/

We study the representations of the simple affine vertex algebras at non-admissible level arising from rank one 4D SCFTs. In particular, we classify the irreducible highest weight modules of $L_{-2}(G_2)$ and $L_{-2}(B_3)$. It is known by the works of Adamovi\'{c} and Per\v{s}e that these vertex algebras can be conformally embedded into $L_{-2}(D_4)$. We also compute the associated variety of $L_{-2}(G_2)$, and show that it is the orbifold of the associated variety of $L_{-2}(D_4)$ by the symmetric group of degree 3 which is the Dynkin diagram automorphism group of $D_4$. This provides a new interesting example of associated variety satisfying a number of conjectures in the context of orbifold vertex algebras., Comment: 29 pages