Last Articles
On the Usefulness of Imperfect models : The Mercator Redemption
A small talk (in french) I did for Human Talks. I deals with imperfect and how we can use them to the full, by bringing the example of infamous Mercator map projection
Robust Learning of a depth map for obstacle avoidance with a monocular stabilized flying camera
This Post presents my PhD Thesis, defended in June 2019
Learning Structure from Motion from motion
Following Depthnet, we present a way to train it with self-supervision, with an algorithm introduced in the paper “Learning Structure-from-motion from motion”
Introducing DepthNet
We present DepthNet, the convolutional neural netowrk introduced in our paper named “End-to-end depth from motion with stabilized monocular videos”