Over the last few years, Artificial Intelligence (AI) and particularly deep learning methods had a transformative impact in fields such as natural language processing or computer vision, significantly advancing the state-of-the-art in problems like parsing natural language, classifying unstructured data, or semantically segmenting images. These same techniques can also empower a next generation of Geographical Information Systems (GIS), providing the ability to combine spatial analysis with fast and near human-level perception, this way facilitating location-based discovery and analysis of relevant information. This is particularly important given the exponentially growing amount of online data, including different online sources of multimedia contents encoding spatial information.

Several recent studies have for instance shown that AI techniques can be used in a variety of geospatial applications, with examples in remote sensing for Earth observation (e.g., segmenting, classifying, downscaling, or fusing ground-level or aerial/satellite imagery), spatial data analysis (e.g., interpolation of geospatial data with generative adversarial networks), or geographical text analysis (e.g., address geo-coding or geo-referencing place references in documents), among many others (e.g., processing historical maps, gazetteer conflation, etc.). Together, these previous research efforts have given birth to a new and interdisciplinary area, often referred to as GeoAI (Geospatial Artificial Intelligence).

Many of the aforementioned applications can clearly highlight important challenges in AI for processing geospatial data, including small sample sizes to inform supervised learning, relative to the complexity of the problems, the lack of ground truth information, the high degree of noise and uncertainty, or the need for appropriately combining human-provided data (e.g., through crowd-sourcing) with automated approaches. Despite many successful applications, overcoming the aforementioned challenges requires additional developments, so that AI techniques can be more widely and easily applicable to a broader range of geospatial applications.

Topics of Interest

The EPIA’21 thematic track on geospatial artificial intelligence is thus concerned with interdisciplinary research related to different aspects of combining AI and Machine Learning (ML) with GIScience. Topics of interest include, but are not limited to:

• Remote Sensing and Earth Observation
• Geospatial Data Science
• Spatial Analysis and Processing of Structured and Semi-Structured Information
• Text Geoparsing and Handling Geographical Information Encoded in Textual Sources
• Exploring Ground-Level Imagery for Proximate Sensing
• Data Conflation and Quality Assurance
• Reasoning with Geospatial Data and Knowledge Bases
• Geospatial Information Retrieval of Multimedia Contents
• Human-Centered Geospatial Artificial Intelligence
• Location-Based Augmented Reality Applications
• Combining Machine Learning and Crowdsourcing Efforts
• Application Areas for Geospatial Artificial Intelligence (e.g., Precision Agriculture, Geospatial Health, etc.)
• AI/ML Challenges and Geospatial Artificial Intelligence (e.g., Explainability, Causality, Efficiency, etc.)

Organizing Committee

Bruno Martins, Instituto Superior Técnico, University of Lisbon (General Organization Chair)
Carlos Damásio, Faculdade de Ciências e Tecnologia, New University of Lisbon (Program Committee Chair)
Jacinto Estima, IADE - Universidade Europeia, Lisboa, Portugal (Program Committee Chair)
João Moura Pires, Faculdade de Ciências e Tecnologia, New University of Lisbon (General Organization Chair)

Program Committee 

Ana Afonso, Faculdade de Ciências, University of Lisbon, Portugal
Ana Navarro, University of Lisbon, Portugal
Ana Teodoro, University of Porto, Portugal
Bruno Martins, Instituto Superior Técnico, University of Lisbon, Portugal
Carlos Damásio, Faculdade de Ciências e Tecnologia, New University of Lisbon, Portugal
Carsten Keßler, Aalborg University, Copenhagen, Sweden
Célia Gouveia, Instituto Português do Mar e da Atmosfera, Portugal
Christian Sallaberry, University of Pau, France
Cidália Fonte, University of Coimbra, Portugal
Devis Tuia, Ecole Polytechnique Fédérale de Lausanne (EPFL), Switzerland
Diego Seco, University of Concepción, Chile
Dino Ienco, INRAE, Montpellier, France
Ekaterina Egorova, University of California at Santa Barbara, United States
Fernando Bação, NOVA IMS, New University of Lisbon, Portugal
Fernando Birra, Faculdade de Ciências e Tecnologia, New University of Lisbon, Portugal
Francisco Javier Lopez-Pellicer, Universidad Zaragoza, Spain
Francisco Javier Zarazaga-Soria, University of Zaragoza, Spain
Gil Gonçalves, University of Coimbra, Portugal
Isabel Trigo, Instituto Português do Mar e da Atmosfera, Portugal
Jacinto Estima, IADE - Universidade Europeia, Lisboa, Portugal
Jérôme Gensel, University of Grenoble, France
João Catalão, Faculdade de Ciências, University of Lisbon, Portugal
João Moura Pires, Faculdade de Ciências e Tecnologia, New University of Lisbon, Portugal
João Valente, Wageningen University & Research, Netherlands
Karine Ferreira, Instituto Nacional de Pesquisas Espaciais, Brasil
Ludovic Moncla, INSA Lyon, France
Ludwig Krippahl, Faculdade de Ciências e Tecnologia, New University of Lisbon, Portugal
Marco Painho, NOVA IMS, New University of Lisbon, Portugal
Maribel Santos, University of Minho, Portugal
Mário Caetano, NOVA IMS, New University of Lisbon, Portugal
Marlène Villanova-Oliver, University of Grenoble, France
Mathieu Roche, University of Montpellier, France
Miguel Luaces, University of A Coruña, Spain
Nuno Datia, Instituto Superior de Engenharia de Lisboa, Portugal
Rafel Santos, Instituto Nacional de Pesquisas Espaciais, Brasil
Rainer Simon, AIT Austrian Institute of Technology, Austria
Roberto Henriques, NOVA IMS, New University of Lisbon, Portugal
Ross Purves, University of Zurich, Switzerland
Sébastien Lefèvre, University of Bretagne-Sud, Vannes, France
Sérgio Freire, European Joint Research Center, Belgium
Stefano De Sabbata, University of Leicester, United Kingdom
Steven Schockaert, Cardiff University, United Kingdom
Susana Nascimento, New University of Lisbon, Portugal
Thomas Corpetti, CNRS, Rennes, France
Yingjie Hu, Department of Geography, University at Buffalo, United States