Train self-supervised vision transformers on overhead imagery with Amazon SageMaker
AWS Machine Learning Blog
AUGUST 16, 2023
We start by downloading the dataset from the terminal of our SageMaker notebook instance: wget [link] tar -xvf BigEarthNet-S2-v1.0.tar.gz Additionally, each folder contains a JSON file with the image metadata. We store the BigEarthNet-S2 images and metadata file in an S3 bucket. The dataset has a size of about 109 GB.
Let's personalize your content