Funded projects in 2023

Multimodal assessment of the depth of sedation of severely ill patients in intensive care unit

Other relevant disciplines: Mathematics/Statistics, Engineering, Medicine, Neuroscience

Abstract to be updated soon.

Key words: anesthesia, signal processing, symbolic representations, physiological signals

Project coordinator: Laurent Oudre (Université Paris Cité)

Semantic coherence integration for optimizing imaging retrieval systems in radiology

Other relevant disciplines: Medicine

Abstract to be updated soon.

Key words: medical imaging, computer vision, content-based image retrieval, deep learning, feature extraction, metric or ranking learning

Project coordinator: Florence Cloppet (Université Paris Cité)

Deep Learning-based EEG Epilepsy Detection and Analysis

Other relevant disciplines: Medicine

Electroencephalogram (EEG) is one of the most common and essential medical signal collected by neural scientists for the analysis of nerve diseases. With the rapid development of medical instruments and data collection techniques, EEG analysis has also been witnessed a dramatic progress. One important problem of EEG analysis is epilepsy pattern detection and analysis. Epilepsy is a brain disease generally associated with seizures, deteriorating the life quality of many patients. This internship targets to design effective deployment schemes of modern deep learning techniques on EEG Epilepsy detection, with a focus on real-world applications for neural scientists.

Key words: EEG analysis, epileptic pattern detection, deep learning, active learning

Project coordinator: Qitong Wang (Université Paris Cité)

Mining molecular dynamics open data

Other relevant disciplines: Computer Science, Linguistics

Abstract to be updated soon.

Key words: RNA modeling, force-field optimization, biomolecular simulations

Project coordinator: Pierre Poulain (Université Paris Cité)

Prediction of protein-carbohydrate binding sites using deep learning methods

Other relevant disciplines: Computer Science, Mathematics/Statistics, Chemistry, Bioinformatics

Abstract to be updated soon.

Key words: protein-carbohydrate interactions, deep learning, structural bioinformatics

Project coordinator: Tatiana Galochkina (Université Paris Cité)

OpenStreetMap and Sentinel-2 data for the production of environmental indices for demographic studies

Other relevant disciplines: Demography

Abstract to be updated soon.

Key words: Remote sensing, Demography, Deep learning, Sentinel 2, OpenStreetMap, Local climate zones, Africa

Project coordinator: Sylvain Lobry (Université Paris Cité)

Deep learning to model genetic pleiotropy to understand the human genetic architecture

Other relevant disciplines: Mathematics/Statistics, Biology

Abstract to be updated soon.

Key words: Pleiotropy, Deep Learning, Convolutional neural network

Project coordinator: Marie Verbanck (Université Paris Cité)

Generalization of a method enabling to update vineyard geographic databases from satellite data

Other relevant disciplines: Earth Sciences/Geosciences

Abstract to be updated soon.

Key words: image time series analysis, deep learning, optical satellite imagery, agriculture monitoring, crop type mapping, vineyard, VENUS images

Project coordinator: Camille Kurtz (Université Paris Cité)

Diffusion Models Based Unpaired Image-to-Image Translation to Reveal Subtle Phenotypes

Other relevant disciplines: Mathematics/Statistics, Biology, neurodevelopment

Unpaired image-to-image translation methods aim at learning a mapping of images from a source domain to a target domain.Recently, these methods proved to be very useful in biological applications to display subtle phenotypic cell variations otherwise invisible to the human eye. However, current models require a large number of images to be trained, while most microscopy experiments remain limited in the number of images they can produce. In this work, we present an improved CycleGAN architecture that employs self-supervised discriminators to alleviate the need for numerous images. We demonstrate quantitatively and qualitatively that the proposed approach outperforms the CycleGAN baseline, including when it is combined with differentiable augmentations. We also provide results obtained with small biological datasets on obvious and non-obvious cell phenotype variations, demonstrating a straightforward application of this method.

Key words: Image-to-image translation, Deep generative models, Diffusion models, Subtle Phenotypes, Neurodevelopment

Project coordinator: Valérie Mezger (Université Paris Cité)

Enhancing earthquake location with domain adapation

Other relevant disciplines: Mathematics/Statistics

Abstract to be updated soon.

Key words: Domain adaptation, machine and deep learning, volcano-seismology, earthquake catalogs

Project coordinator: Léonard Seydoux (IPGP)

Exploration of press articles related to Covid-19 at the European level within the Covid-19 Museum

Other relevant disciplines: Computer Science, Mathematics/Statistics, Linguistics

Abstract to be updated soon.

Key words: Covid-19 Museum, Newspaper analysis, Timeline

Project coordinator: Yves Rozenholc (Université Paris Cité)

Machine Learning for Photometric redshift estimation of LSST galaxies

Other relevant disciplines: Computer Science, Mathematics/Statistics

Abstract to be updated soon.

Key words: machine learning, convolutional networks, astrophysics

Project coordinator: Simona Mei (Université Paris Cité)