Planetary Feature Fields are Scalable Earth Representations
cs.CV, cs.LG
Submitted: 2026-09-29
Updated: 2026-09-29
Comments: 28 pages, 16 figures, 7 tables
License: http://creativecommons.org/licenses/by/4.0/
The gist: Satellite observations, precomputed embeddings, and map products describe the same evolving Earth, yet are stored as independent, petabyte-scale data products.
Terminology
Abstract
Satellite observations, precomputed embeddings, and map products describe the same evolving Earth, yet are stored as independent, petabyte-scale data products. Their continued growth calls for compact representations of multiple products while preserving spatial and temporal detail. We introduce Planetary Feature Fields (PFFs), which exploit redundancy across data products by modeling them jointly as continuous functions of space and time at planetary scale. PFFs are spatially local explicit-implicit (hybrid) neural fields. Each field shares a factored feature volume---a decomposition of an explicit 3D grid with smaller factors---across products, while lightweight implicit decoders reconstruct individual products across multiple timesteps. PFFs reconstruct EO products over space and time more accurately than single-product fields at matched compression rates. At 1800 times compression relative to the uncompressed source data, reconstructed features retain approximately 90% or more of the performance achieved with the original features on pixel-level segmentation, change detection, and patch-level classification tasks. PFFs can add new timesteps by extending their factored feature volumes and add new products by attaching new decoders, while leaving existing outputs unchanged. PFFs reduce end-to-end feature access latency by an order of magnitude relative to evaluated API and cloud-storage pipelines.
Sources
- Spatial Functa: Scaling Functa to ImageNet Classification and Generation
- AlphaEarth Foundations: An embedding field model for accurate and efficient global mapping from sparse label data
- MIND the Gap: A Geographic Implicit Neural Representation with Adjustable Spatial Scale
- TerraCodec: Compressing Optical Earth Observation Data
- Global and Dense Embeddings of Earth: Major TOM Floating in the Latent Space
- SLED: Scalable Location Encoding via Distillation
- CD-NGP: A Fast Scalable Continual Representation for Dynamic Scenes
- Lightweight, Pre-trained Transformers for Remote Sensing Timeseries
- Better Together: Evaluating the Complementarity of Earth Embedding Models
Related papers
- Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
- AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
- Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
- MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
- TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
- A Survey on Efficient Vision-Language-Action Models