(M/F) Postdoc Position in Machine learning for Underwater Acoustics and Source Localisation

Mobilize mathematical and computer methods to solve a theoretical problem relating to the simulation of a model;
Optimize its programming on a target machine and offer the appropriate tools.
Provide researchers in a field with expertise in the use of mathematical methods and computer techniques for the modeling and simulation of a physical phenomenon
Orient the choice on the relevant methods and tools according to the problem posed and the architecture of the targeted computing machines
Design methods for the modeling, calculation and visualization of results
Evaluate the quality of the codes, the quality of the results and their interpretation
Participate in national and international research projects and associated publications

• Task 1 : method development
• 1.1 : solving direct problem using AI (replace the propagation model with a trained neural network) : Solving an inverse problem in ASM with ML. Our efforts will focus on situations where ML has obvious added value (monitoring of fluctuations in the water column and / or the detection / localization of sources in a specific area of interest. The candidate will work on data from the ALMA experimental area and / or with experimental area "Shallow Water 2006 "..). The learning will be done also on simulations.

• 1.2 : solving inverse problem using AI (replace the inverse algorithm by a trained neural network) : The objective is to replace the numerical evaluation of the direct problem (in our case a model of propagation) by a method of ML of the regression type. Interest of this approach is to accelerate the resolution of the inversion problem (estimation of the environment and / or location of sources) by at least an order of magnitude.

• Task 2 : Experimental data analysis : validate the proposed method on experimental at-sea data, notably using the DGA ALMA dataset. : Database resulting from the ALMA (Acoustic Laboratory for Marine Applications) measurement campaigns at sea can be used. The various experiments provide access to a wide range of variable environments (bathymetry, type of antenna, identification of phenomena linked to the fluctuation of the environment and impact on antenna treatments or even source localization in passive acoustics).

This proposed job is a postdoctoral position at GIPSA-Lab Grenoble. The overall topic is the use of Artificial Intelligence (AI) for Anti-Submarine Warfare (ASW).

Practically, the main mission will be to revisit classical ocean acoustics inverse problems (source localization and environmental inversion) using IA. The performance of the proposed IA methods will be quantified and compared to state-of-the-art methods (notably Bayesian inversion methods). Performance metric will include accuracy, uncertainty, but also computation time and robustness to mismatch (both environmental and statistical mismatch).

Candidates selected during the CNRS recruitment procedures must be validated by the DGA prior to the signing of the employment contract between the candidate and the CNRS.

Keywords: Detection / localization / classification and environmental characterization for underwater acoustics characterization in the era of Big Data: Contribution and opportunity of Artificial Intelligence.

Gipsa-lab is a CNRS research unit joint, Grenoble-INP (Grenoble Institute of Technology), University of Grenoble under agreement with Inria, Observatory of Sciences of the Universe of Grenoble. With 350 people including about 150 doctoral students, Gipsa-lab is a multidisciplinary research unit developing both basic and applied researches on complex signals and systems. Gipsa-lab develops projects in the strategic areas of energy, environment, communication, intelligent systems, life and health and linguistic engineering.
Thanks to the research activities, Gipsa-lab maintains a constant link with the economic environment through a strong partnership with companies. Gipsa-lab staff is involved in teaching and training in the various universities and engineering schools of the Grenoble academic area (Université Grenoble Alpes). Gipsa-lab is internationally recognized for the research achieved in Automatic & Diagnostics, Signal Image Information Data Sciences, Speech and Cognition. The research unit develops projects in 16 teams organized in 4 Reseach centers
.Automatic & Diagnostic
.Data Science
.Geometry, Learning, Information and Algorithms
.Speech -cognition
Gipsa-lab regroups 150 permanent staff and around 250 no-permanent staff (Phd, post-dotoral students, visiting scholars, trainees in master…)
The SIGMAPHY (Signal-Image-Physics) team develops advanced signal/image processing methods for specific applications where an understanding of the underlying physics is required. The main research topics are
• signal and wave propagation
• remote sensing from planes and satellites
• transient signals analysis