Show all
Show all
Title Description Supervisors Status
Beyond Gaussians with Diagonal Covariance How much information do deep neural network weights contain after training? One way of answering this question is modelling the … Konstantinos Pitas Available
How flat is my minimum? Flat minima are claimed to have links with good generalization in deep neural networks [1]. However such links to good … Konstantinos Pitas Available
Better Uncertainty Estimates for DNNs v2! Deriving uncertainty estimates for feedforward DNN predictions is critical in a number of tasks. Unfortunately the outputs of the softmax … Konstantinos Pitas Available
Development, assessment and optimization of an Active Nonlinear Electroacoustic Resonator Acoustic metamaterials (AMMs) and metasurfaces (AMSs) are engineered structures aimed at achieving unprecedented acoustic properties not available in nature. They … Xinxin Guo Available
Room modes identification and characterization and its potential application Room modes can significantly impair the quality of sound diffusion, in particular at low-frequencies and in small auditoria. These low-frequency … Thach Pham Vu Available
Development of a microphone array to track mobile noise sources Microphone arrays can be used to locate and track moving sound sources, thanks to advanced signal processing techniques. Such techniques … Hervé Lissek Available
Multi-actuated panel for active sound absorption The distributed mode loudspeaker (DML) technology is a flat panel loudspeaker from which sound is generated a set of electromechanical … Hervé Lissek Available
Implementation of a low-cost real-time reverberation algorithm The LTS2-Acoustics is working on binaural audio synthesis for applications in hearing aid technologies. Our activities deal with the improvement … Vincent Grimaldi Available
Data-driven research on graph neural networks I am looking for motivated students that wish to work on the development of deep learning algorithms for graph data. … Andreas Loukas Available
VENOVA : an actively modified cylindrical cross-saxophone After some research, modeling, computations, adjustments and a bit of luck, the team of engineers from Yamaha have developed a … Thomas Laurence Available
In-Situ Assessment of Active Electroacoustic Absorbers in Rooms at Low-Frequencies Room modes can significantly impair the quality of sound diffusion, in particular at low-frequencies and in small auditoria. These low-frequency … Thach Pham Vu Available
Development of a remote quiet zone demo In many workspaces, low-frequency noises (LFN), as experienced with industrial machinery, electric transformer and air flow machinery, are the most … Romain Boulandet Available
Design of a new flat loudspeaker concept The electrodynamic moving-coil loudspeaker, as patented by Kellogg and Rice in 1924, is the most spread electroacoustic diffuser in the … Thomas Laurence Available
Development of an active acoustic metasurface Metamaterials are engineered materials with a subwavelength structure that enables extraordinary propagation and manipulation of waves. In acoustics, they can … Thomas Laurence Available
Implementation of an active sourdine for wind instruments Electroacoustic resonators are devices comprising (at least) one electrodynamic loudspeaker and an electronic instrumentation (feedback based on pressure/velocity sensing, shunt … Hervé Lissek, Etienne Rivet, Romain Boulandet Available
Sound absorption of 3D-printed membranes The 3D printer technology represent now a nainstream tool for designing engineering tools that were not (or almost not) possible … Hervé Lissek Available
Protein reconstruction from multiple images with Deep Learning Single-particle cryo-electron microscopy (cryo-EM) is a Nobel-prized technology that aims to characterize the 3D structure of proteins at the atomic … Michaël Defferrard, Laurène Donati Assigned
Detection of fake news in sites and online networks There are numerous sources of information on the net. They can be more or less biaised or reliable and it … Benjamin Ricaud Assigned
Geometric Deep Learning for Climate and Weather Modeling Climate change is one of the grand challenges facing humanity in the 21st century. Humankind can mainly act on two … Michaël Defferrard Finished
Learning on manifolds In many applications, data lie on non-Euclidean manifolds. Examples on the sphere include planetary data (such as temperature, wind, aerial … Michaël Defferrard, Martino Milani Finished
An empirical study of rotation equivariant neural networks The goal of the project is to empirically study the difference in performance of Neural Networks that are either equivariant … Michaël Defferrard Finished
Assessment of Low-rank approximation techniques in sound field reconstruction Room modes can significantly impair the quality of sound diffusion, in particular at low-frequencies and in small auditoria. In practice, … Thach Pham Vu Finished
Better uncertainty estimates for DNNs DNNs are increasingly used in sensitive environments such as healthcare. In these sensitive environments a number of important questions arise. … Konstantinos Pitas Finished
Geometric Deep Learning for Fluid Dynamics Use Machine Learning on graphs and manifolds to solve problems in computational fluid dynamics. Michaël Defferrard Finished
Music outside the head In nature and most daily life situations, sounds of the world are naturally perceived as coming from outside the head. … Vincent Grimaldi Finished