Main content
VibrantVS: A High-Resolution Vision Transformer for Forest Canopy Height Estimation
- Tony Chang
- Kiarie Ndegwa
- Andreas Gros
- Vincent A. Landau
- Luke J. Zachmann
- Bogdan State
- Mitchell A. Gritts
- Colton W. Miller
- Nathan E. Rutenbeck
- Scott Conway
- Guy Bayes
Date created: | Last Updated:
: DOI | ARK
Creating DOI. Please wait...
Category: Project
Description: This project is a collection of the test dataset for reproducing the analysis from the VibrantVS model manuscript. The test dataset contains paired NAIP, aerial lidar CHM, and VibrantVS CHM inference. All data have been spatially intersected and temporally intersected within one year. Data is generated at 0.5m pixel resolution representing a 500m x 500m footprint (1000 x 1000 pixels). All data is projected in EPSG:6931[WGS 84 / NSIDC EASE-Grid 2.0 North]. Within the S3 bucket containing the data there are a 3 folders 'naip', 'lidar', 'vibrantvs' containing the individual Cloud-Optimized GeoTiffs (COGs) of each sample. For faster access, we have also provided a .zip file of the naip, lidar, and vibrantvs data that contain all the COGs. Additionally, we have provided a `test_data_catalog.gpkg` geopackage that provides the sample unique identifier (tile_id), the correspond s3 URI for each datatype (naip_s3_uri, lidar_s3_uri, vibrantvs_s3_uri), and the corresponding footprint they represent.
Add important information, links, or images here to describe your project.
Files
Files can now be accessed and managed under the Files tab.