Posts by Collection





Scene understanding with 2D and 3D data from UAVs


2D and 3D scene understanding, with application in rail station and railway monitoring (PRI Drosofiles) and 3D modelling and semantic mapping for search-and-rescue (FP7 Inachus Project)

Deep Learning for Remote Sensing


Tutorial at JURSE’2019 by Loïc Landrieu (IGN, French Mapping Agency) and Bertrand Le Saux (ONERA, French Aerospace Agency)

Learning to understand Earth observation images with weak and unreliable ground truth


In this paper we discuss the issues of using inexact and inaccurate ground truth in the context of supervised learning. To leverage large amounts of Earth observation data for training algorithms, one often has to use ground truth which was not been carefully assessed. We address both the problems of training and evaluation. We first propose a weakly supervised approach for training change classifiers which is able to detect pixel-level changes in aerial images. We then propose a data poisoning approach to get a reliable estimate of the accuracy that can be expected from a classifier, even when the only ground-truth available does not match the reality. Both are assessed on practical land use and land cover applications.

Beyond Labels: Weakly-supervised, Continual and Semi-supervised Learning for Earth Observation


Keynote talk at ICPR2021 Pattern Recognition in Remote Sensing PRRS workshop.

Abstract: More and more data (and their corresponding meta-data) have allowed the wide adoption of automatic machine learning approaches for Earth observation. These methods, often relying on supervised learning, are designed (and succeed!) to obtain high performances on numerous and ever larger carefully prepared benchmarks. But what happens when you go in the wild? When you cannot trust the labels, or worse, when no labels exist? Domain adaptation and generalisation issues appear, leading to unpredictable results.

In this talk I will present several approaches which learn beyond labels. First, I will present a weak supervision method which allows to train a neural network model when labels are inadequate or noisy. Second, I will talk about continual learning for adapting models with the help of a human-in-the-loop. Finally, I will address semi-supervised learning, or how to train models from both labelled and unlabelled data. All those approaches pave the way to today’s great challenge of Earth observation: how to develop generic models able to handle the plethora of data now available?