Project Background

GeoNDC emerged from research at Beijing Normal University's Advanced Interdisciplinary Institute of Satellite Applications, addressing a fundamental challenge in earth observation: the exponential growth of satellite data outpacing our ability to store, transmit, and analyze it.

Traditional remote sensing data cubes store raw spectral bands, requiring massive storage infrastructure and limiting accessibility. GeoNDC proposes a paradigm shift: instead of storing what satellites saw, we store how the Earth works. By learning implicit neural representations of planetary vegetation dynamics, we achieve unprecedented compression while enabling instant spatial queries.

The project name reflects our core innovation: compressing global earth observation data into neural network weights that can be "executed" like programs, querying any location instantly without decompression.

Core Team

Jianbo Qi

Lead Researcher

Beijing Normal University
Advanced Interdisciplinary Institute of Satellite Applications

Mengyao Li

Researcher

Beijing Normal University
Satellite Data Compression

Baogui Jiang

Researcher

Beijing Normal University
Neural Network Architecture

Yidan Chen

Researcher

Beijing Normal University
Validation & Testing

Qiao Wang

Researcher

Beijing Normal University
Data Processing Pipeline

Institution

Beijing Normal University

Advanced Interdisciplinary Institute of Satellite Applications

Beijing, China

Contact

For academic collaborations, data inquiries, or general questions:

Email Us GitHub HuggingFace

Response time: 1-3 business days

Collaboration Opportunities

We welcome collaborations in the following areas:

  • Academic research partnerships on neural representation for remote sensing
  • Data providers interested in compressing and sharing earth observation archives
  • Commercial applications requiring high-resolution vegetation data
  • Joint funding applications for satellite data analysis projects

Please reach out via email with a brief introduction of your research or organization.