Hugging Face's logo

embed2scale
/
TerraCodec-1.0-FP-S2L2A

TerraTorch
Compression
Neutral Codecs
Earth Observation
Model card Files
xet
 Community
Isabelle-Wittmann commited on
Commit
b279425
·
1 Parent(s): 2392f74

update readme

Browse files

Signed-off-by: Isabelle Wittmann <isabelle.wittmann1@ibm.com>

Files changed (2) hide show
  1. .gitattributes +1 -0
  2. README.md +4 -7
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ *.png filter=lfs diff=lfs merge=lfs -text
README.md CHANGED
@@ -8,13 +8,11 @@ homepage: https://github.com/IBM/TerraCodec
8
 
9
  **Neural Compression for Earth Observation**
10
 
11
- ![License](https://img.shields.io/badge/License-Apache%202.0-blue)
12
- ![arXiv](https://img.shields.io/badge/arXiv-2510.12670-b31b1b)
13
- ![GitHub](https://img.shields.io/badge/GitHub-IBM%2FTerraCodec-black?logo=github)
14
- ![PyPI](https://img.shields.io/badge/PyPI-terracodec-blue?logo=pypi)
15
 
16
- TerraCodec (TEC) is a family of pretrained neural compression codecs for **multispectral Sentinel-2 satellite imagery**.
17
- The models compress optical Earth observation data using learned latent representations and entropy coding.
18
 
19
  Compared to classical codecs such as JPEG2000 or WebP, TerraCodec achieves **3–10× higher compression at comparable reconstruction quality** on multispectral satellite imagery. Temporal models further improve compression by exploiting redundancy across seasonal image sequences of satellite imagery.
20
 
@@ -70,7 +68,6 @@ Inputs should be standardized per spectral band using dataset statistics. For S2
70
 
71
  ```python
72
  mean = torch.tensor([793.243, 924.863, 1184.553, 1340.936, 1671.402, 2240.082, 2468.412, 2563.243, 2627.704, 2711.071, 2416.714, 1849.625])
73
-
74
  std = torch.tensor([1160.144, 1201.092, 1219.943, 1397.225, 1400.035, 1373.136, 1429.170, 1485.025, 1447.836, 1652.703, 1471.002, 1365.307])
75
  ```
76
 
 
8
 
9
  **Neural Compression for Earth Observation**
10
 
11
+ [![arXiv](https://img.shields.io/badge/arXiv-2510.12670-b31b1b)](https://arxiv.org/abs/2510.12670)
12
+ [![GitHub](https://img.shields.io/badge/GitHub-IBM%2FTerraCodec-black?logo=github)](https://github.com/IBM/TerraCodec)
13
+ [![PyPI](https://img.shields.io/badge/PyPI-terracodec-blue?logo=pypi)](https://pypi.org/project/terracodec/)
 
14
 
15
+ TerraCodec (TEC) is a family of pretrained neural compression codecs for **multispectral Sentinel-2 satellite imagery**. The models compress optical Earth observation data using learned latent representations and entropy coding.
 
16
 
17
  Compared to classical codecs such as JPEG2000 or WebP, TerraCodec achieves **3–10× higher compression at comparable reconstruction quality** on multispectral satellite imagery. Temporal models further improve compression by exploiting redundancy across seasonal image sequences of satellite imagery.
18
 
 
68
 
69
  ```python
70
  mean = torch.tensor([793.243, 924.863, 1184.553, 1340.936, 1671.402, 2240.082, 2468.412, 2563.243, 2627.704, 2711.071, 2416.714, 1849.625])
 
71
  std = torch.tensor([1160.144, 1201.092, 1219.943, 1397.225, 1400.035, 1373.136, 1429.170, 1485.025, 1447.836, 1652.703, 1471.002, 1365.307])
72
  ```
73