TerraCodec: Compressing Optical Earth Observation Data
METADATA ONLY
Loading...
Author / Creator
Date
2027
Publication Type
Conference Paper
ETH Bibliography
yes
Citations
Scopus:
Altmetric
METADATA ONLY
Data
Rights / License
Abstract
Earth observation (EO) satellites produce massive streams of multispectral image time series, posing pressing challenges for storage and transmission. Yet, learned EO compression remains fragmented and lacks publicly available, large-scale pretrained codecs. Moreover, prior work has largely focused on image compression, leaving temporal redundancy and EO video codecs underexplored. To address these gaps, we introduce TerraCodec (TEC), a family of learned codecs pretrained on Sentinel-2 EO data. TEC includes efficient multispectral image variants and a Temporal Transformer model (TEC-TT) that leverages dependencies across time. To overcome the fixed-rate setting of today’s neural codecs, we present Latent Repacking, a novel method for training flexible-rate transformer models that operate on varying rate-distortion settings. TerraCodec outperforms classical codecs, achieving 3–10× higher compression at equivalent image quality. Beyond compression, TEC-TT enables zero-shot cloud inpainting, surpassing state-of-the-art methods on the AllClear benchmark. Our results establish neural codecs as a promising direction for Earth observation. Our code and models are publicly available at https://github.com/IBM/TerraCodec.
Permanent link
Publication status
published
External links
Book title
Machine Learning and Knowledge Discovery in Databases. Research Track
Journal / series
Volume
16943
Pages / Article No.
271 - 287
Publisher
Springer
Event
European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML PKDD 2026)
Edition / version
1st Edition