CopernicusLAC Chile project engineer Gabriela Jerez and scientific advisor Valentín Barriere participated in the 42nd International Conference on Machine Learning (ICML 2025), one of the most important global gatherings in the field of artificial intelligence. The event took place between 13 and 19 July in Vancouver, Canada.

Gabriela Jerez attended the event in person, which featured workshops such as ‘TerraBytes: Towards global datasets and models for Earth observation’, which addressed the current challenges of modelling with satellite data on a global scale. The workshop stood out for its focus on topics such as spatio-temporal biases, the curation of representative datasets by region, and the development of efficient learning techniques, including contrastive, self-supervised, continuous, and domain-adaptive methods.

During his participation, Jerez explored new applications in land use monitoring, biodiversity, fires and floods, and established links with GeoAI and remote sensing researchers. The workshop was organised by SpatialLabs, the team responsible for Copernicus artificial intelligence products.

Valentín Barriere participated remotely as co-organiser of the TerraBytes workshop, leading the team in coordinating the event, supervising the scientific review process, and publishing the accepted papers in a special issue of the Proceedings of Machine Learning Research (PMLR) series.

Participation in ICML 2025 reinforces CopernicusLAC Chile’s commitment to applied research, the strategic use of satellite data, and capacity building in artificial intelligence for Latin America and the Caribbean.