NeonTreeEvaluation Benchmark, CanopyRS Preprocessed Tiles: Individual Tree Crown Boxes on NEON Airborne RGB Imagery

Provider
Weinstein, Marconi and White (weecology); preprocessed by the CanopyRS authors
License
CC-BY-4.0 (source)
Size
10 files
Format
Parquet (in sampled files)
Stored at
Hugging Face (https://huggingface.co/datasets/CanopyRS/NeonTreeEvaluation)
Access
Free and public. No account with the data provider is needed.
DOI
10.5281/zenodo.5914554
Category
Environmental Science
Source
https://huggingface.co/datasets/CanopyRS/NeonTreeEvaluation

About this dataset

A tiled version of the NeonTreeEvaluation benchmark of Weinstein, Marconi and White (weecology), prepared by the CanopyRS authors for the SelvaBox tree crown detection study. The original benchmark supplies National Ecological Observatory Network (NEON) airborne data for forest sites across the United States (RGB imagery, LiDAR, 426-band hyperspectral and canopy height model tiles) with hand-drawn bounding boxes around individual trees, for evaluating tree detection from airborne imagery. This release keeps the RGB imagery and box annotations, cut into tiles in COCO format and split into 912 train, 934 validation and 194 test tiles, as Parquet shards under data/ (about 3.3 GB) loadable with datasets.load_dataset. Useful for training and comparing tree crown detectors across NEON forest sites and as an out-of-distribution test for tropical models. The full original data, including LiDAR and hyperspectral tiles, is on Zenodo (DOI 10.5281/zenodo.5914554). Original data licensed CC BY 4.0; cite Weinstein et al. and the SelvaBox paper.

How to load this dataset in Python

from datasets import load_dataset
ds = load_dataset("CanopyRS/NeonTreeEvaluation")
print(ds)

Data structure

File types in sampled files: Parquet

Files

Example file: data/test-00000-of-00001.parquet

Analyze this dataset without downloading it

On Scigantic, NeonTreeEvaluation Benchmark, CanopyRS Preprocessed Tiles: Individual Tree Crown Boxes on NEON Airborne RGB Imagery opens in a hosted JupyterLab notebook with the data already mounted, so you can read files where they are stored instead of copying them first. Each notebook includes an AI assistant that can see the mounted files and write and run analysis code. Open the dataset page to launch a notebook, or browse the full catalog.

This page is also available as Markdown and as JSON.

Related Environmental Science datasets