Early 2021, I gave a series of talks on the topic of Beyond Labels: Weakly-supervised, Continual and Semi-supervised Learning for Earth Observation, in the following venues.
- [Jan. 2021, 10] Keynote at ICPR’2020-2021/PRRS workshop: Slides.
- [Jan. 2021, 14] Talk at the Dutch AI Technical Session organised by Iris Reimerink from Netherlands’ Kadaster, with a focus on change detection.
- [Fev. 2021, 25] Seminar at Ireland’s Centre for Applied AI, CeADAR Seminar series, with extension to energy-based models.
The material for these presentations in mostly based on the works of Javiera Castillo-Navarro, Rodrigo Caye Daudt, Gaston Lenczner and Veda Sunkara. Further information can be found in the following papers:
Multitask learning for large-scale semantic change detection R. Caye Daudt, B. Le Saux, A. Boulch, Y. Gousseau Computer Vision and Image Understanding (CVIU), vol. 187, Oct. 2019
Guided Anisotropic Diffusion and Iterative Learning for Weakly Supervised Change Detection R. Caye Daudt, B. Le Saux, A. Boulch, Y. Gousseau IEEE/CVF Conf. on Computer Vision and Pattern Recognition (CVPR / EarthVision 19) Long Beach, Cal., USA, June 2019
DISIR: Deep image segmentation with interactive refinement G. Lenczner, B. Le Saux, N. Luminari, A. Chan-Hon-Tong, G. Le Besnerais, ISPRS Annals, Virtual / Nice, France, August 2020
Interactive Learning for Semantic Segmentation in Earth Observation G. Lenczner, B. Le Saux, N. Luminari, A. Chan-Hon-Tong, G. Le Besnerais, ECML/PKDD MACLEAN, Virtual / Ghent, Belgium, September 2020
Semi-Supervised Semantic Segmentation in Earth Observation: The MiniFrance Suite, Dataset Analysis and Multi-task Network Study J. Castillo-Navarro, B. Le Saux, A. Boulch, N. Audebert, S. Lefèbre, Maching Learning, accepted, to appear.