Prithvi-EO-2.0: A Versatile Multi-Temporal Foundation Model for Earth Observation Applications

2024-12-03

arXiv.org

AI Summary

Weak spatiotemporal modeling capability and poor cross-resolution generalization hinder remote sensing foundation models. To address these limitations, this paper introduces Prithvi-EO-2.0—the first multimodal remote sensing foundation model integrating both temporal and geospatial coordinate embeddings. Designed via a task-driven paradigm with continuous domain-expert involvement in architecture customization, it is pre-trained on a harmonized global dataset of 4.2 million Landsat–Sentinel-2 time-series samples at 30 m resolution. Prithvi-EO-2.0 supports dual-scale Transformer architectures (300M and 600M parameters). On GEO-Bench, it achieves an 8% improvement over its predecessor and outperforms six state-of-the-art remote sensing foundation models across all benchmarks. Notably, it demonstrates unprecedented wide-resolution generalization—from 0.1 m to 15 m—enabling substantial gains in accuracy and robustness for downstream applications including disaster response, crop mapping, and environmental monitoring.

Technology Category

• Computer Vision: Remote Sensing / Geospatial AI
• Machine Learning: Deep Neural Architectures and Foundation Models
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Security and Privacy: Data transparency and provenance

Abstract

This technical report presents Prithvi-EO-2.0, a new geospatial foundation model that offers significant improvements over its predecessor, Prithvi-EO-1.0. Trained on 4.2M global time series samples from NASA's Harmonized Landsat and Sentinel-2 data archive at 30m resolution, the new 300M and 600M parameter models incorporate temporal and location embeddings for enhanced performance across various geospatial tasks. Through extensive benchmarking with GEO-Bench, the 600M version outperforms the previous Prithvi-EO model by 8% across a range of tasks. It also outperforms six other geospatial foundation models when benchmarked on remote sensing tasks from different domains and resolutions (i.e. from 0.1m to 15m). The results demonstrate the versatility of the model in both classical earth observation and high-resolution applications. Early involvement of end-users and subject matter experts (SMEs) are among the key factors that contributed to the project's success. In particular, SME involvement allowed for constant feedback on model and dataset design, as well as successful customization for diverse SME-led applications in disaster response, land use and crop mapping, and ecosystem dynamics monitoring. Prithvi-EO-2.0 is available on Hugging Face and IBM terratorch, with additional resources on GitHub. The project exemplifies the Trusted Open Science approach embraced by all involved organizations.

Problem

• Earth Observation
• Disaster Response
• Environmental Monitoring

Innovation

• Earth Observation Model
• Iterative Development
• Open Science Principles