Current Work (2019-2023) : 3D humans

I am currently finishing a PhD with the MORPHEO team of the INRIA on the representation and synthesis of 3D humans.

During my PhD, I investigated deep learning architecture to efficiently encode dense 3D motion in compact latent representations. In a first work, focusing on a structured latent space with generative properties. And in a second work, a more generic latent representation of long sequences that was leveraged and showed state-of-the-art result on motion completion tasks.

I also studied the task of motion retargeting which aims at transferring the motion of a source character to a target character to directly animate a character with the raw output of a multi-view aquisition system.

Older work (2019) : 2D computer vision

During a 6 months master thesis at Cranfield University, I explored the use of data-driven models to perform facial segmentation of farm animals (pigs). The objective being to identify a pig given a 2D RGB video/image as input.

Older work (2018): 3D printing

I made a 3 months internship with the MFX team at the LORIA. During this internship, I was task with investigating the use of textures generated using Gabor Noise to automatically infer the printing head pathing in 3D printers.