Project description
Helping experts to discover and analyse patterns
Today’s machine learning algorithms can provide predictions powerful enough to sometimes replace human interpretation. The ERC-funded DISCOVER project, however, will aim instead to develop new approaches to assist experts in identifying and analysing patterns. Specifically, it will take as input images without any annotation, automatically identify consistent patterns and model their variation and evolution so that an expert can more easily analyse them. To this end, the project will develop two approaches: one based on analysing correspondences and the other on learning interpretable image models. The methods will be developed in two application domains: historical documents and Earth imagery.
Objective
The goal of this project is to shift the dominant paradigm of learning-based computer vision: instead of systems attempting to replace human interpretation by providing predictions, we will develop approaches to assist experts in identifying and analyzing patterns. Indeed, while the success of deep learning on visual data is undeniable, applications are often limited to the supervised learning scenario where the algorithm tries to infer a label for a new image based on the annotations made by experts in a reference dataset. In contrast, we will take as input images without any annotation, automatically identify consistent patterns and model their variation and evolution, so that an expert can more easily analyze them.
I will introduce and develop the concept of visual structures. Their key features will be their interpretability, in terms of correspondences, deformations, or properties of the observed images, and their ability to incorporate prior knowledge about the data and expert feedback. I propose two complementary approaches to formally define and identify visual structures: one based on analyzing correspondences, the other on learning interpretable image models.
We will develop visual structures in two domains in which breakthrough progress will open up new scientific discoveries: historical documents and Earth imagery. For example, from temporal series of multispectral Earth images, we will identify types of moving objects, areas with different types of vegetation or constructions, and model the evolution of their characteristics, which may correspond to changes in their activity or life cycle. Ultimately, experts will still be needed to select relevant visual structures and perform analysis, but DISCOVER will revolutionize their work, trivializing tedious annotation tasks and even allowing them to work on issues they would have been hard-pressed to identify in the raw data.
Fields of science
Programme(s)
- HORIZON.1.1 - European Research Council (ERC) Main Programme
Topic(s)
Funding Scheme
ERC - Support for frontier research (ERC)Host institution
77455 Marne La Vallee Cedex 2
France