-
NeurIPS 2024 ML4CFD Competition: Harnessing Machine Learning for Computational Fluid Dynamics in Airfoil Design
Authors:
Mouadh Yagoubi,
David Danan,
Milad Leyli-abadi,
Jean-Patrick Brunet,
Jocelyn Ahmed Mazari,
Florent Bonnet,
maroua gmati,
Asma Farjallah,
Paola Cinnella,
Patrick Gallinari,
Marc Schoenauer
Abstract:
The integration of machine learning (ML) techniques for addressing intricate physics problems is increasingly recognized as a promising avenue for expediting simulations. However, assessing ML-derived physical models poses a significant challenge for their adoption within industrial contexts. This competition is designed to promote the development of innovative ML approaches for tackling physical…
▽ More
The integration of machine learning (ML) techniques for addressing intricate physics problems is increasingly recognized as a promising avenue for expediting simulations. However, assessing ML-derived physical models poses a significant challenge for their adoption within industrial contexts. This competition is designed to promote the development of innovative ML approaches for tackling physical challenges, leveraging our recently introduced unified evaluation framework known as Learning Industrial Physical Simulations (LIPS). Building upon the preliminary edition held from November 2023 to March 2024, this iteration centers on a task fundamental to a well-established physical application: airfoil design simulation, utilizing our proposed AirfRANS dataset. The competition evaluates solutions based on various criteria encompassing ML accuracy, computational efficiency, Out-Of-Distribution performance, and adherence to physical principles. Notably, this competition represents a pioneering effort in exploring ML-driven surrogate methods aimed at optimizing the trade-off between computational efficiency and accuracy in physical simulations. Hosted on the Codabench platform, the competition offers online training and evaluation for all participating solutions.
△ Less
Submitted 30 June, 2024;
originally announced July 2024.
-
ML4PhySim : Machine Learning for Physical Simulations Challenge (The airfoil design)
Authors:
Mouadh Yagoubi,
Milad Leyli-Abadi,
David Danan,
Jean-Patrick Brunet,
Jocelyn Ahmed Mazari,
Florent Bonnet,
Asma Farjallah,
Marc Schoenauer,
Patrick Gallinari
Abstract:
The use of machine learning (ML) techniques to solve complex physical problems has been considered recently as a promising approach. However, the evaluation of such learned physical models remains an important issue for industrial use. The aim of this competition is to encourage the development of new ML techniques to solve physical problems using a unified evaluation framework proposed recently,…
▽ More
The use of machine learning (ML) techniques to solve complex physical problems has been considered recently as a promising approach. However, the evaluation of such learned physical models remains an important issue for industrial use. The aim of this competition is to encourage the development of new ML techniques to solve physical problems using a unified evaluation framework proposed recently, called Learning Industrial Physical Simulations (LIPS). We propose learning a task representing a well-known physical use case: the airfoil design simulation, using a dataset called AirfRANS. The global score calculated for each submitted solution is based on three main categories of criteria covering different aspects, namely: ML-related, Out-Of-Distribution, and physical compliance criteria. To the best of our knowledge, this is the first competition addressing the use of ML-based surrogate approaches to improve the trade-off computational cost/accuracy of physical simulation.The competition is hosted by the Codabench platform with online training and evaluation of all submitted solutions.
△ Less
Submitted 3 March, 2024;
originally announced March 2024.
-
Wireless Power Transfer in Space using Flexible, Lightweight, Coherent Arrays
Authors:
Alex Ayling,
Austin Fikes,
Oren S. Mizrahi,
Ailec Wu,
Raha Riazati,
Jesse Brunet,
Behrooz Abiri,
Florian Bohn,
Matan Gal-Katziri,
Mohammed Reza M. Hashemi,
Sharmila Padmanabhan,
Damon Russell,
Ali Hajimiri
Abstract:
Space solar power (SSP), envisioned for decades as a solution for continuous, stable, and dynamically dispatchable clean energy, has seen tremendous interest and a number of experimental demonstrations in the last few years. A practical implementation has been elusive to date, owing to the high launch costs associated with heavy, rigid photovoltaic (PV) and wireless power transfer (WPT) arrays. Li…
▽ More
Space solar power (SSP), envisioned for decades as a solution for continuous, stable, and dynamically dispatchable clean energy, has seen tremendous interest and a number of experimental demonstrations in the last few years. A practical implementation has been elusive to date, owing to the high launch costs associated with heavy, rigid photovoltaic (PV) and wireless power transfer (WPT) arrays. Lightweight and flexible solutions for WPT have been demonstrated terrestrially but, to date, have not been deployed and tested in space. In this paper, we present an experimental space demonstration of a lightweight, flexible WPT array powered by custom radio frequency integrated circuits (RFICs). The transmit arrays, receive arrays, and the rest of the system were operated and tested for eight months in Low Earth Orbit (LEO). Results from these experiments, including pointing of the array's beam to Earth and its detection by a ground station, are presented and discussed in detail. Observations and results from this mission uncover existing strengths and weaknesses that inform future steps toward realizing SSP.
△ Less
Submitted 16 February, 2024; v1 submitted 26 January, 2024;
originally announced January 2024.
-
Eikonal phase retrieval: Unleashing the fourth generation sources potential for enhanced propagation based tomography on biological samples
Authors:
Alessandro Mirone,
Joseph Brunet,
Leandre Admans,
Renaud Boistel,
Morgane Sowinski,
Camille Berruyer,
Henri Payno,
Elodie Boller,
Pierre Paleo,
Claire L. Walsh,
Peter D. Lee,
Paul Tafforeau
Abstract:
The evolution of synchrotrons towards higher brilliance beams has increased the possible sample-to-detector propagation distances for which the source confusion circle does not lead to geometrical blurring. This makes it possible to push near-field propagation driven phase contrast enhancement to the limit, revealing low contrast features which would otherwise remain hidden under an excessive nois…
▽ More
The evolution of synchrotrons towards higher brilliance beams has increased the possible sample-to-detector propagation distances for which the source confusion circle does not lead to geometrical blurring. This makes it possible to push near-field propagation driven phase contrast enhancement to the limit, revealing low contrast features which would otherwise remain hidden under an excessive noise. Until today this possibility was hindered, in many objects of scientific interest, by the simultaneous presence of strong phase gradient regions and low contrast features. The strong gradients, when enhanced with the now possible long propagation distances, induce such strong phase effects that the linearisation assumptions of current state-of-the-art single-distance phase retrieval filters are broken, and the resulting image quality is jeopardized. Here, we introduce a new iterative phase-retrieval algorithm and compare it with the Paganin phase-retrieval algorithm, in both the monochromatic and polychromatic cases, obtaining superior image quality. In the polychromatic case the comparison was done with an extrapolated Paganin algorithm obtained by reintroducing, into our phase retrieval algorithm, the linearization approximations underlying the Paganin forward model. Our work provides an innovative algorithm which efficiently performs the phase retrieval task over the entire near-field range, producing images of exceptional quality for mixed attenuation objects.
△ Less
Submitted 18 June, 2024; v1 submitted 22 November, 2023;
originally announced November 2023.
-
Deep Learning for Vascular Segmentation and Applications in Phase Contrast Tomography Imaging
Authors:
Ekin Yagis,
Shahab Aslani,
Yashvardhan Jain,
Yang Zhou,
Shahrokh Rahmani,
Joseph Brunet,
Alexandre Bellier,
Christopher Werlein,
Maximilian Ackermann,
Danny Jonigk,
Paul Tafforeau,
Peter D Lee,
Claire Walsh
Abstract:
Automated blood vessel segmentation is vital for biomedical imaging, as vessel changes indicate many pathologies. Still, precise segmentation is difficult due to the complexity of vascular structures, anatomical variations across patients, the scarcity of annotated public datasets, and the quality of images. We present a thorough literature review, highlighting the state of machine learning techni…
▽ More
Automated blood vessel segmentation is vital for biomedical imaging, as vessel changes indicate many pathologies. Still, precise segmentation is difficult due to the complexity of vascular structures, anatomical variations across patients, the scarcity of annotated public datasets, and the quality of images. We present a thorough literature review, highlighting the state of machine learning techniques across diverse organs. Our goal is to provide a foundation on the topic and identify a robust baseline model for application to vascular segmentation in a new imaging modality, Hierarchical Phase Contrast Tomography (HiP CT). Introduced in 2020 at the European Synchrotron Radiation Facility, HiP CT enables 3D imaging of complete organs at an unprecedented resolution of ca. 20mm per voxel, with the capability for localized zooms in selected regions down to 1mm per voxel without sectioning. We have created a training dataset with double annotator validated vascular data from three kidneys imaged with HiP CT in the context of the Human Organ Atlas Project. Finally, utilising the nnU Net model, we conduct experiments to assess the models performance on both familiar and unseen samples, employing vessel specific metrics. Our results show that while segmentations yielded reasonably high scores such as clDice values ranging from 0.82 to 0.88, certain errors persisted. Large vessels that collapsed due to the lack of hydrostatic pressure (HiP CT is an ex vivo technique) were segmented poorly. Moreover, decreased connectivity in finer vessels and higher segmentation errors at vessel boundaries were observed. Such errors obstruct the understanding of the structures by interrupting vascular tree connectivity. Through our review and outputs, we aim to set a benchmark for subsequent model evaluations using various modalities, especially with the HiP CT imaging database.
△ Less
Submitted 22 November, 2023;
originally announced November 2023.
-
SOniCS: Develop intuition on biomechanical systems through interactive error controlled simulations
Authors:
Arnaud Mazier,
Sidaty El Hadramy,
Jean-Nicolas Brunet,
Jack S. Hale,
Stéphane Cotin,
Stéphane P. A. Bordas
Abstract:
This new approach allows the user to experiment with model choices easily and quickly without requiring in-depth expertise, as constitutive models can be modified by one line of code only. This ease in building new models makes SOniCS ideal to develop surrogate, reduced order models and to train machine learning algorithms for uncertainty quantification or to enable patient-specific simulations. S…
▽ More
This new approach allows the user to experiment with model choices easily and quickly without requiring in-depth expertise, as constitutive models can be modified by one line of code only. This ease in building new models makes SOniCS ideal to develop surrogate, reduced order models and to train machine learning algorithms for uncertainty quantification or to enable patient-specific simulations. SOniCS is thus not only a tool that facilitates the development of surgical training simulations but also, and perhaps more importantly, paves the way to increase the intuition of users or otherwise non-intuitive behaviors of (bio)mechanical systems. The plugin uses new developments of the FEniCSx project enabling automatic generation with FFCx of finite element tensors such as the local residual vector and Jacobian matrix. We validate our approach with numerical simulations such as manufactured solutions, cantilever beams, and benchmarks provided by FEBio. We reach machine precision accuracy and demonstrate the use of the plugin for a real-time haptic simulation involving a surgical tool controlled by the user in contact with a hyperelastic liver. We include complete examples showing the use of our plugin for simulations involving Saint Venant-Kirchhoff, Neo-Hookean, Mooney-Rivlin, and Holzapfel Ogden anisotropic models as supplementary material.
△ Less
Submitted 24 August, 2022;
originally announced August 2022.
