EXPLAINABLE AI PIPELINES FOR BIG COPERNICUS DATA

Informations projet

DeepCube

N° de convention de subvention: 101004188

DOI 10.3030/101004188

Projet clôturé

Date de signature de la CE 27 Octobre 2020

Date de début 1 Janvier 2021

Date de fin 31 Decembre 2023

Financé au titre de

• INDUSTRIAL LEADERSHIP - Leadership in enabling and industrial technologies – Space

Coût total € 3 999 687,50

Contribution de l’UE € 3 999 687,50

Coordonné par

ETHNIKO ASTEROSKOPEIO ATHINON
Greece

CORDIS fournit des liens vers les livrables publics et les publications des projets HORIZON.

Les liens vers les livrables et les publications des projets du 7e PC, ainsi que les liens vers certains types de résultats spécifiques tels que les jeux de données et les logiciels, sont récupérés dynamiquement sur OpenAIRE .

Livrables

Demonstrators, pilots, prototypes (5)

Creation of training datasets, data cubes & ontologies - v2

This is an update scaleup of the v1 deliverable

DeepCube Platform - v1

Platform integration and system

DeepCube Platform - v2

Updated version of based on changes made to DeepCube’s platform. In addition, this deliverable includes the user manual and documentation of the platform.

Analytics and DL architectures for fire risk assessment UC-v1

Report containing a consolidate version of all DL architectures that DeepCubes pilots will implement

Analytics and DL architectures for the droughts UC- v1

Report containing a consolidate version of all DL architectures that DeepCubes pilots will implement

Documents, reports (12)

Final dissemination and communication report

Detailed report of all activities (M01-M36) that took place during the course of the project.

Project Management & Quality plan

Guidelines to ensure high quality research development and reporting along with a project time plan

DeepCube platform requirements, specs and architecture-v2

Updated version of the technical requirements deliverable based on any updates from 1st Use Case implementation and evaluation cycle

EO and non-EO data ingestion report - v2

This is an update of the v1 deliverable.

DeepCube technical components-v1

Inputs on developments adjustments finetuning of the technological components

DeepCube technical components - v2

Updates from 1st Use Case implementation and evaluation cycle This deliverable also includes user manuals and documentations for the individual components

Status of Liaison activities v2

Updates on liaison activities

Initial Communication and Dissemination plan

At the early stage of the proposal to update communication plan and dissemination plan identifying target audiences key messages channels tools and metrics

DeepCube platform requirements, specs and architecture-v1

Technical specifications and the architectural design of the DeepCube platform

EO and non-EO data ingestion report - v1

Information on all data ingested along with precise information on data origin

Mid-term dissemination plan

Report updates on dissemination and communication plan for the next period M18M36 including also a detailed summary of the main activities that took place during the first 18 months

Status of Liaison activities v1

Provision of tangible liaison activities with other projects

Data sets, microdata, etc (1)

Creation of training datasets, data cubes & ontologies-v1

Describe in detail both the process and the generated trained datasets. Report all technical specifications and stored datasets stored on all Data Cubes. This report provides the ontologies and mappings to be used for realizing the DeepCube Semantic Cube. The deliverable provides the data cubes with ARD per se, to allow further exploitation.

Open Research Data Pilot (2)

Data Management Plan v2

Updated version of the DMP report

Data Management Plan v1

DMP formulation contributing in making data FAIR

Websites, patent fillings, videos etc. (1)

Website & Material

Website up and running along with any other material will support outreach activities brochures leaflets video newsletter etc

Publications

Conference proceedings (11)

Pluto: A global volcanic activity early warning system powered by large scale self-supervised deep learning on InSAR data

Auteurs: Nikolaos Ioannis Bountos, Dimitrios Michail, Themistocles Herekakis, Angeliki Thanasou, Ioannis Papoutsis
Publié dans: EGU General Assembly 2023, 2023
Éditeur: European Geosciences Union
DOI: 10.5194/egusphere-egu23-5913

Learning drivers of climate-induced human migrations with Gaussian processes

Auteurs: Jose M. Tarraga, Maria Piles, Gustau Camps-Valls
Publié dans: NeurIPS 2020 Workshop on Machine Learning for the Developing World, 2020
Éditeur: NeurIPS 2020 Workshop on Machine Learning for the Developing World

Sen4AgriNet: A Harmonized Multi-Country, Multi-Temporal Benchmark Dataset for Agricultural Earth Observation Machine Learning Applications

Auteurs: D. Sykas, I. Papoutsis, D. Zografakis
Publié dans: 2021 IEEE International Geoscience and Remote Sensing Symposium IGARSS, Numéro 12 October 2021, 2021, Page(s) 5830-5833
Éditeur: IEEE
DOI: 10.1109/igarss47720.2021.9553603

Mesogeos: A multi-purpose dataset for data-driven wildfire modeling in the Mediterranean

Auteurs: Spyros Kondylatos, Ioannis Prapas, Gustau Camps-Valls, Ioannis Papoutsis
Publié dans: 37th Conference on Neural Information Processing Systems (NeurIPS 2023) Track on Datasets and Benchmarks, 2023
Éditeur: NeurIPS
DOI: 10.48550/arxiv.2306.05144

Deep Learning Methods for Daily Wildfire Danger Forecasting

Auteurs: Ioannis Prapas, Spyros Kondylatos, Ioannis Papoutsis, Gustau Camps-Valls, Michele Ronco, Miguel-Ángel Fernández-Torres, Maria Piles Guillem, Nuno Carvalhais
Publié dans: Workshop on Artificial Intelligence for Humanitarian Assistance and Disaster Response, 35th Conference on Neural Information Processing Systems (NeurIPS 2021), Numéro 4 November 2021, 2021
Éditeur: NeurIPS

Inspecting the link between climate and human displacement with Explainable AI and Causal inference

Auteurs: José María Tárraga Habas, Michele Ronco, Maria Teresa Miranda, Eva Sevillano Marco, Qiang Wang, María Piles, Jordi Muñoz, and Gustau Camps-Valls
Publié dans: EGU General Assembly 2022, 2022
Éditeur: EGU22-11200
DOI: 10.5194/egusphere-egu22-11200

Explainable deep learning for wildfire danger estimation

Auteurs: Ronco, M., Prapas, I., Kondylatos, S., Papoutsis, I., Camps-Valls, G., Fernández-Torres, M.-Á., Piles Guillem, M., and Carvalhais, N.
Publié dans: EGU General Assembly 2022, 2022
Éditeur: EGU22-11787
DOI: 10.5194/egusphere-egu22-11787

AutoAblation: Automated Parallel Ablation Studies for Deep Learning

Auteurs: Sina Sheikholeslami, Moritz Meister, Tianze Wang, Amir H. Payberah, Vladimir Vlassov, Jim Dowling
Publié dans: EuroMLSys '21: Proceedings of the 1st Workshop on Machine Learning and Systems, 2021, Page(s) 55-61
Éditeur: ACM
DOI: 10.1145/3437984.3458834

DEEPCUBE: EXPLAINABLE AI PIPELINES FOR BIG COPERNICUS DATA

Auteurs: Ioannis Papoutsis, Alkyoni Baglatzi, Souzana Touloumtzi, Markus Reichstein, Nuno Carvalhais, Fabian Gans, Gustau Camps-Valls, Maria Piles, Theofilos Kakantousis, Jim Dowling, Manolis Koubarakis, Dimitris Bilidas, Despina-Athanasia Pantazi, George Stamoulis, Christophe Demange, Leo-Gad Journel, Marco Bianchi, Chiara Gervasi, Alessio Rucci, Ioannis Tsampoulatidis, Eleni Kamateri, Tarek Habib, Alejan
Publié dans: Proceedings of the 2021 conference on Big Data from Space (BiDS’21), 2021
Éditeur: Proceedings of the 2021 conference on Big Data from Space (BiDS’21)

Hephaestus: A large scale multitask dataset towards InSAR understanding

Auteurs: N.I. Bountos, I. Papoutsis, D. Michail, A. Karavias, P. Elias, I. Parcharidis
Publié dans: 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), 2022
Éditeur: IEEE
DOI: 10.48550/arxiv.2204.09435

Learning to forecast vegetation greenness at fine resolution over Africa with ConvLSTMs

Auteurs: Claire Robin, Christian Requena-Mesa, Vitus Benson, Lazaro Alonso, Jeran Poehls, Nuno Carvalhais, Markus Reichstein
Publié dans: Tackling Climate Change with Machine Learning: workshop at NeurIPS 2022, 2022
Éditeur: NeurIPS 2022
DOI: 10.48550/arxiv.2210.13648

Peer reviewed articles (10)

Assessing the Added Value of Sentinel-1 PolSAR Data for Crop Classification

Auteurs: Maria Ioannidou; Alkiviadis Koukos; Vasileios Sitokonstantinou; Ioannis Papoutsis; Charalampos Kontoes
Publié dans: Remote Sensing; Volume 14; Numéro 22; Pages: 5739, Numéro 13 November 2022, 2022, ISSN 2072-4292
Éditeur: Multidisciplinary Digital Publishing Institute (MDPI)
DOI: 10.3390/rs14225739

Integration of a Deep-Learning-Based Fire Model Into a Global Land Surface Model

Auteurs: Rackhun Son, Tobias Stacke, Veronika Gayler, Julia E. M. S. Nabel, Reiner Schnur, Lazaro Alonso, Christian Requena-Mesa, Alexander J. Winkler, Stijn Hantson, Sönke Zaehle, Ulrich Weber, Nuno Carvalhais
Publié dans: Journal of Advances in Modeling Earth Systems, Numéro Volume 16, Numéro1, 2024, ISSN 1942-2466
Éditeur: American Geophysical Union
DOI: 10.1029/2023ms003710

Self-supervised Contrastive Learning for Volcanic Unrest Detection

Auteurs: Nikolaos Ioannis Bountos, Ioannis Papoutsis, Dimitrios Michail, Nantheera Anantrasirichai
Publié dans: IEEE Geoscience and Remote Sensing Letters, Numéro Volume 19, 2021, Page(s) 1-5, ISSN 1558-0571
Éditeur: IEEE
DOI: 10.1109/lgrs.2021.3104506

A Sentinel-2 Multiyear, Multicountry Benchmark Dataset for Crop Classification and Segmentation With Deep Learning

Auteurs: Dimitrios Sykas; Maria Sdraka; Dimitrios Zografakis; Ioannis Papoutsis
Publié dans: IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, Numéro Volume 15, 2022, ISSN 2151-1535
Éditeur: IEEE
DOI: 10.1109/jstars.2022.3164771

Toward Robust Parameterizations in Ecosystem-Level Photosynthesis Models

Auteurs: Shanning Bao, Lazaro Alonso, Siyuan Wang, Johannes Gensheimer, Ranit De, Nuno Carvalhais
Publié dans: Journal of Advances in Modeling Earth Systems, Numéro Volume 15, Numéro 8, 2023, ISSN 1942-2466
Éditeur: American Geophysical Union
DOI: 10.1029/2022ms003464

Exploring interactions between socioeconomic context and natural hazards on human population displacement

Auteurs: Michele Ronco, José María Tárraga, Jordi Muñoz, María Piles, Eva Sevillano Marco, Qiang Wang, Maria Teresa Miranda Espinosa, Sylvain Ponserre, Gustau Camps-Valls
Publié dans: Nature Communications, Numéro 14, 2023, ISSN 2041-1723
Éditeur: Nature Publishing Group
DOI: 10.1038/s41467-023-43809-8

Role of locality, fidelity and symmetry regularization in learning explainable representations

Auteurs: Michele Ronco, Gustau Camps-Valls
Publié dans: Neurocomputing, Numéro Volume 562, 2023, ISSN 0925-2312
Éditeur: Elsevier BV
DOI: 10.1016/j.neucom.2023.126884

Wildfire Danger Prediction and Understanding With Deep Learning

Auteurs: S. Kondylatos, I. Prapas, M. Ronco, I. Papoutsis, G. Camps-Valls, M. Piles, M. Fernández-Torres, N. Carvalhais
Publié dans: Geophysical Research Letters, Numéro Volume 49, Numéro17, 2022, ISSN 1944-8007
Éditeur: American Geophysical Union
DOI: 10.1029/2022gl099368

Learning From Synthetic InSAR With Vision Transformers: The Case of Volcanic Unrest Detection

Auteurs: Nikolaos Ioannis Bountos; Dimitrios Michail; Ioannis Papoutsis
Publié dans: IEEE Transactions on Geoscience and Remote Sensing (Volume 60), Numéro 08 June 2022, 2022, ISSN 1558-0644
Éditeur: IEEE
DOI: 10.1109/tgrs.2022.3180891

Benchmarking and scaling of deep learning models for land cover image classification

Auteurs: Ioannis Papoutsis, Nikolaos Ioannis Bountos, Angelos Zavras, Dimitrios Michail, Christos Tryfonopoulos
Publié dans: ISPRS Journal of Photogrammetry and Remote Sensing, Numéro Volume 195, January 2023, 2023, ISSN 1872-8235
Éditeur: Elsevier
DOI: 10.1016/j.isprsjprs.2022.11.012

Other (2)

Learning class prototypes from Synthetic InSAR with Vision Transformers

Auteurs: Nikolaos Ioannis Bountos, Dimitrios Michail, Ioannis Papoutsis
Publié dans: 2022
Éditeur: arXiv
DOI: 10.48550/arxiv.2201.03016

Efficient deep learning models for land cover image classification

Auteurs: Ioannis Papoutsis, Nikolaos-Ioannis Bountos, Angelos Zavras, Dimitrios Michail, Christos Tryfonopoulos
Publié dans: 2021
Éditeur: arXiv