EarthView: A Large Scale Remote Sensing Dataset for Self-Supervision

Diego Velazquez, Pau Rodriguez, Sergio Alonso, Josep M. Gonfaus, Jordi Gonzalez, Gerardo Richarte, Javier Marin, Yoshua Bengio, Alexandre Lacoste; Proceedings of the Winter Conference on Applications of Computer Vision (WACV) Workshops, 2025, pp. 1228-1237

This paper presents EarthView a comprehensive dataset specifically designed for self-supervision on remote sensing data intended to enhance deep learning applications on Earth monitoring tasks. The dataset spans 15 tera pixels of global remote-sensing data combining imagery from a diverse range of sources including NEON Sentinel and a novel release of 1m spatial resolution data from Satellogic. Our dataset provides a wide spectrum of image data with varying resolutions harnessed from different sensors and organized coherently into an accessible Hugging Face dataset in parquet format. This data spans five years from 2017 to 2022. Accompanying the dataset we introduce EarthMAE a tailored Masked Autoencoder developed to tackle the distinct challenges of remote sensing data. Trained in a self-supervised fashion EarthMAE effectively processes different data modalities such as hyper-spectral multi-spectral topographical data segmentation maps and temporal structure. We regard this innovative combination of an expansive diverse dataset and a versatile model adapted for self-supervised learning as a significant stride forward in deep learning for Earth monitoring.