About Me

I am a postdoctoral researcher fellow at McGill University and Mila (Quebec Artificial Intelligence Institute), in collaboration with David Rolnick. My research projects are focused on multimodal and multitask deep learning applied to forest monitoring. I’m also a core team member of Climate Change AI leading the webinar team.

I completed my Ph.D. in March 2022 in collaboration between Institut Polytechnique de Paris (Télécom Paris; IDS department, IMAGES Team) and valeo.ai, under the supervision of Patrick Pérez, Florence Tupin and Alasdair Newson. The aim of my Ph.D. work was to use and adapt deep neural network architectures for scene understanding using automotive radar data and multi-sensor fusion.

I graduated from M.Sc. Machine Learning and Big Data at Télécom Paris in 2018, M.Sc. (1 and 2) in Statistical Modelling at Paris Panthéon Sorbonne University in 2016 and B.S. in Applied Mathematics at Paris Diderot University in 2014.

During my Ph.D., I co-organized the Deep Learning Working Group of the IMAGES Team (Télécom Paris). Presentations are available here.

News

• 12/2023: Our new work FoMo-Bench: a multi-modal, multi-scale and multi-task Forest Monitoring Benchmark for remote sensing foundation models is now on ArXiv! Code and datasets will be release soon.
  
  
• 11/2023: Our new work OpenForest: A data catalogue for machine learning in forest monitoring is now on ArXiv! The OpenForest catalogue is also available on github.
  
  
• 02/2023: Starting a core team member position at Climate Change AI as the webinar team leader.
  
  
• 09/2022: Starting my postdoc on forest monitoring using deep learning at McGill and Mila in Montreal, Canada.
  
  
• 04/2022: My 2021 ICPR and ICCV articles have been quoted at the Nvidia GTC by Geoffrey Bouquot, group CTO and VP at Valeo. Link of the presentation is here.
  
  
• 03/2022: Finally defended my PhD, the manuscript is available here.
  
  
• 03/2022: Our recent work ``Raw High-Definition Radar for Multi-Task Learning” has been accepted at CVPR 2022!
  
  
• 12/2021: New preprint available on ArXiv: ``Raw High-Definition Radar for Multi-Task Learning”. The new RADIal dataset and the code of our proposed deep learning model are available on the github repo!
  
  
• 08/2021: Our work on multi-view radar semantic segmentation has been accepted at ICCV 2021!
  
  
• 07/2021: Last but not least, the RAD tensors of the CARRADA dataset are now available.
  
  
• 05/2021: Code for Multi-View Radar Semantic Segmentation (MVRSS) and pretrained models have been publicly released on github here.
  
  
• 04/2021: A new version of CARRADA is available! Have a look to the repo here.
  
  
• 04/2021: New preprint available on ArXiv: Multi-View Radar Semantic Segmentation.
  
  
• 01/2021: The paper of CARRADA has been presented in a poster session of ICPR 2020.
  
  
• 10/2020: First paper accepted to ICPR 2020!