We provide scripts to prepare datasets for training, including PointOdyssey, TartanAir, Spring, and Waymo. For evaluation, we also provide a script for preparing the Sintel dataset.
Note
The scripts provided here are for reference only. Please ensure you have obtained the necessary licenses from the original dataset providers before proceeding.
To download the pre-trained models, run the following commands:
cd data
wget https://download.europe.naverlabs.com/ComputerVision/DUSt3R/DUSt3R_ViTLarge_BaseDecoder_512_dpt.pth -P ../checkpoints/
cd ..
To download and prepare the PointOdyssey dataset, execute:
cd data
bash download_pointodyssey.sh
cd ..
To download and prepare the TartanAir dataset, execute:
cd data
bash download_tartanair.sh
cd ..
To download and prepare the Spring dataset, execute:
cd data
bash download_spring.sh
cd ..
To download and prepare the Waymo dataset, follow these steps:
- Set up Google Cloud SDK (if you haven't done so already):
curl https://sdk.cloud.google.com | bash
exec -l $SHELL
gcloud init
gcloud auth login
- Download the Waymo dataset:
cd data
bash download_waymo.sh
cd ..
- Preprocess the dataset and create training pairs:
python datasets_preprocess/preprocess_waymo.py
python datasets_preprocess/waymo_make_pairs.py
To download and prepare the Sintel dataset for evaluation, execute:
cd data
bash download_sintel.sh
cd ..