metadata
license: cc-by-nc-4.0
dataset_info:
features:
- name: input
sequence:
sequence:
sequence: float32
- name: label
dtype: float32
splits:
- name: train
num_bytes: 621854296128
num_examples: 9949032
- name: validation
num_bytes: 167623664728
num_examples: 2681807
- name: test
num_bytes: 206495839888
num_examples: 3303722
download_size: 368286700835
dataset_size: 995973800744
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*
🌲 AGBD: A Global-scale Biomass Dataset 🌳
Authors: Ghjulia Sialelli (gsialelli@ethz.ch), Torben Peters, Jan Wegner, Konrad Schindler
Usage
You can use this dataset with the following lines of code:
#!pip install datasets
from datasets import load_dataset
dataset = load_dataset("prs-eth/AGBD", streaming=True)["train"] # or test, val
This will load an IterableDataset — you can learn more about how to operate on them at this link.
25x25 Version
This dataset is the bigger 25x25 Version if you want to have access to the smaller 15x15 patches goto huggingface.co/datasets/prs-eth/AGBD_15
Dataset Description
Each dataset sample consists of a pair of pre-cropped, pre-normalized images and their corresponding biomass labels. The full project page including links to preprocessed uncropped data can be found on github.com/ghjuliasialelli/AGBD/
Image Details
Each image contains 24 channels, organized as follows:
- Spectral Bands: B01, B02, B03, B04, B05, B06, B07, B08, B8A, B09, B11, B12
- Geographical Coordinates: lat_cos, lat_sin, lon_cos, lon_sin
- ALOS PALSAR Bands: alos_hh, alos_hv
- Canopy Heights: ch, ch_std
- Land Cover Information: lc_cos, lc_sin, lc_prob
- Digital Elevation Model: dem
Channel Structure
(B01 B02 B03 B04 B05 B06 B07 B08 B8A B09 B11 B12) | (lat_cos, lat_sin, lon_cos, lon_sin) | (alos_hh, alos_hv) | (ch, ch_std) | (lc_cos, lc_sin, lc_prob) | dem