IBM, HuggingFace, and NASA Open-Sources Watsonx․ai Foundation Model: NASA’s First Openly Available AI Foundation Model and the Largest Geospatial Model on HuggingFace

IBM and the open-source AI platform Hugging Face have jointly announced the release of the watsonx.ai geospatial foundation model. This remarkable AI model, developed using NASA’s satellite data, represents a significant advancement in climate science and Earth research. The primary objective of this partnership is to democratize access to AI and foster accelerated innovations in these crucial areas.

The realm of climate science faces a pressing challenge in accessing the most up-to-date data, given the ever-changing environmental conditions. Despite a growing influx of data—anticipated to reach 250,000 terabytes from new missions by 2024—analyzing these extensive datasets remains a formidable task for scientists and researchers. To address this concern, IBM developed an AI foundation model for geospatial data earlier this year as part of a Space Act Agreement with NASA.

By making the geospatial foundation model available on Hugging Face, a prominent proponent of open-source AI models, the collaboration aims to facilitate greater cooperation and information sharing within the AI community. This move is anticipated to expedite the development of impactful solutions that benefit the planet.

The geospatial foundation model was jointly trained on Harmonized Landsat Sentinel-2 satellite data (HLS) across the continental United States over a year. The model has exhibited an impressive 15 percent enhancement over existing techniques while requiring only half the labeled data. The model can be adapted with further fine-tuning for various tasks, including deforestation tracking, crop yield prediction, and greenhouse gas detection and monitoring. IBM and NASA also collaborate with Clark University to explore applications such as time-series segmentation and similarity research.

IBM’s geospatial model harnesses its foundation model technology, part of the company’s broader endeavor to create and train AI models for diverse tasks, leveraging knowledge transfer between scenarios. In July, IBM launched Watsonx, an AI and data platform empowering enterprises to scale and accelerate the impact of advanced AI using reliable data. A commercial version of the geospatial model, integrated into the IBM Environmental Intelligence Suite (EIS), is scheduled for release later this year.

In conclusion, the partnership between IBM and Hugging Face, fortified by NASA’s satellite data, represents a promising opportunity to drive scientific progress and deepen our understanding of the planet’s climate. The open-source nature of the model is poised to empower researchers and scientists worldwide in tackling pressing environmental challenges.

Check out the Reference Article and Project Page. All Credit For This Research Goes To the Researchers on This Project. Also, don’t forget to join our 27k+ ML SubReddit, 40k+ Facebook Community, Discord Channel, and Email Newsletter, where we share the latest AI research news, cool AI projects, and more.