MillionST

MillionST MillionST is a large-scale satellite image time series dataset curated for pre-training spatiotemporal foundation models for Earth observation. The dataset contains approximately 1 million satellite images from 100,000 geographic locations, with each location observed across 10 temporal phases over five years. It is designed to capture diverse geospatial changes, seasonal variations, and long-term land-surface dynamics. MillionST was introduced in the paper: TiMo:… See the full description on the dataset page:

Source: Hugging Face Hub (Hillui/MillionST). Metadata imported from the dataset’s Hub tags.