BiDS'23 Signup

Julia's High-Performance Geospatial tools: From Maps to Insights.

Lazaro Alonso (Max-Planck Institute for Biogeochemistry)

Showcases of using Julia's plotting and geospatial analysis ecosystem for the generation of interactive dashboards and high level visualisation for generating insights into Earth system data.

The Importance of Seeding in Open-Source

Marco Bernasocchi (OPENGIS.ch GmbH)

Seeding is crucial in open-source software development. It provides a solid foundation for projects, enabling effective contributions. Seed projects outline architecture, coding standards, and best practices, fostering collaboration and consistency. They also serve as learning resources for newcomers. This talk will delve into how seeding empowers open-source development, attracting more contributors and improving code quality.

Efficient ground truth thanks to QField

Marco Bernasocchi (OPENGIS.ch GmbH)

Explore the power of QField, an open-source field mapping app, in our lightning talk "Efficient Ground Truth Thanks to QField". Discover its user-friendly interface, offline capabilities, and seamless QGIS integration. Learn how QField enhances fieldwork efficiency, improves data accuracy, and boosts productivity through real-world examples. Transform your field data collection process with QField.

Scivision: making computer vision more accessible to the earth observation community

Alejandro Coca-Castro (The Alan Turing Institute)

Scivision (https://sci.vision) is an open-source software tool, an open catalogue of datasets and models, and a community of computer vision experts and users. Scivision aims to accelerate scientific computer vision across domains by sharing and matching models and datasets through the Scivision catalogue.

Your Open Source resources platform - OSS4gEO

Codrina Maria Ilie (Terrasigna)

OSS4gEO is a community led initiative part of the wider Open Innovation framework at ESA that works to develop an open, interactive, user intuitive platform for a constantly updated, comprehensive and detailed overview of the dynamic environment of the open source digital infrastructure for geospatial data storage, processing and visualisation systems.

ISME-HYDRO - an Information information infrastructure for sustainable monitoring and exploitation of water resources of dams and rivers based on Earth observation and AI

Mariana Damova (Mozaika)

Water resources management requires daily monitoring of hydrological and environmental parameters in order to evaluate the status of the water bodies they are in charge with. They need a wholesome view of different kinds of data and forecasts for hydrological and hydrodynamic features. We show how using meteorological satellite data and several AI methods give a viable solution for water resources management and beyond.

The OECD Municipal Atlas

Claudia Baranzelli (OECD)

As municipalities worldwide strive to make informed decisions for sustainable development and better quality of life, there is a growing need for comprehensive tools that facilitate the visualisation and analysis of key indicators at the local level. The OECD Geospatial Lab is designing a tool to meet this demand, which allows users to interact with a diverse range of local indicators, integrating data from different sources, from official statistics to unconventional geospatial data sources.

The Beauty of Seeing Beyond the Visible: Where On-Board AI Meets Hyperspectral Imaging for (Not Only) Soil Analysis

Jakub Nalepa (KP Labs)

Bringing AI on board imaging satellites, such as Intuition-1 by KP Labs which will be in orbit in a couple of days, offers enormous scalability of such solutions. Hear the story of the HYPERVIEW Challenge which attracted almost 160 teams from around the world wanting to build robust soil analysis techniques for the better planet. Wanna do more & see beyond the visible? Join us at https://platform.ai4eo.eu/seeing-beyond-the-visible-permanent and push the state of the art of soil analysis.

DGGS and Xarray: A draft implementation

Justus Magin (CNRS)

Discrete Global Grid Systems (DGGS) are an innovative form of global spatial reference systems. DGGS typically discretize the Earth's surface into a regular grid via a hierarchical tesselation with unique cell identifiers. During this conference, we implemented an experimental package for Xarray that provides the necessary operations to work with DGGS-encoded large-scale Earth Observation data.

The COALA ENABLING PLATFORM FOR DIGITAL FARMING

Francesco Vuolo (Boku)

The COALA enabling platform is an automated Sentinel-2 processing infrastructure and a front- end Application Programming Interface (API) that provides value-added products for irrigation and crop nutrient management. The COALA enabling platform ensures high quality and accuracy in the derived information while offering enhanced efficiency, scalability, customization, and cost savings. I will introduce some examples and provide a token for application and testing

The Humanitarian Data Cube: EO Data for Humanitarian Operations

Valentin Pesendorfer (UN World Food Programme)

Earth Observation data is a key source in informing humanitarian operations at different levels through different organizations and stakeholders. Yet, EO data products geared towards humanitarian use remain scarce and hard to use. The Humanitarian Data Cube (HDC) is a cloud native platform developed by WFP to process EO data and serve analysis-ready EO data for its operations and the wider humanitarian community.

Diffusion Model for generating rare objects on satellite data

Aurélien Lac (Thales)

We worked on a finetuning of Stable diffusion for image generation from text of satellite imagery. We can also guide generation from geometric condition. This can be used to do knowledge transfer to panoptic segmentation with open vocabulary on satellite imagery.

Overview of vision-language models applied remote sensing data

Christian Rakotondrainibe (Thales Services Numériques)

Our primary goal is to show the immense potential of vision-language models in the domain of earth observation, unlocking new avenues for enhanced data interpretation and insights.

We conducted a meticulous analysis of state-of-the-art models like CLIP, ALBEF or BLIP and fine-tuned them on remote sensing datasets. We then analysed them diverse tasks including visual question answering, caption generation, image retrieval and visual grounding.

Lightning Talks Schedule