Skip to content

Latest commit

 

History

History
43 lines (29 loc) · 2.24 KB

README.md

File metadata and controls

43 lines (29 loc) · 2.24 KB

SynthCave PyTest

This repository contains the code for the dataset generation, as well as the code for the models used in the SynthCave paper. The models are implemented in PyTorch and include a CNN, ASTGCN, TSViT, and PSTNet.

View Paper

About

SynthCave is a synthetic dataset for 3D odometry estimation in cave-like environments. It contains synthetic LiDAR data in three different forms: point clouds, depth-images, and graphs, along with IMU and ground-truth data.

The raw data was recorded with simulated LiDAR and IMU sensors, using the Minecraft Measurement Mod with Minecaft version 1.20.2. The dataset is publicly available on Kaggle.

Installation

  1. Install requirements:

    pip install torch==2.0.1 --index-url https://download.pytorch.org/whl/cu118
pip install -r requirements.txt

  2. Compile PointNet++ CUDA layers for PSTNet:

    cd baselines/model/PSTNet/modules/pointnet2_ops_lib
python setup.py install

    PSTNet is the only model that requires CUDA and cannot be run on CPU.

  3. Test installation:

    pytest baselines/model
pytest synthcave

  4. (Optional) Manually test the forward pass of every model:

    • CNN: python baselines/model/CNN/CNN.py
    • ASTGCN: python baselines/model/ASTGCN/ASTGCN.py
    • TSViT: python baselines/model/TSViTcls/TSViT.py
    • PSTNet: python baselines/model/PSTNet/PSTNet.py