Hi everyone,
As an employee of Robovision (https://robovision.ai/) I followed the Jetson AI Specialist course.
As the final project assignment, I developed an AI project on a Jetson device, which I would like to showcase here.
I’ve integrated a 3D point cloud segmentation network (RandlaNet: [1911.11236] RandLA-Net: Efficient Semantic Segmentation of Large-Scale Point Clouds) in a docker container which can run on Jetson. The implementation of RandlaNet was based on GitHub - aRI0U/RandLA-Net-pytorch: PyTorch implementation of RandLA-Net.
The network is wrapped in an easy-to-use UI tool to showcase capturing, labeling and prediction in realtime. To capture 3D data I’ve integrated with the intel realsense L515.
The source code for this project can be found here:
A demonstration video is available on youtube here:
I hope this was interesting for you or that you could get some inspiration.
Feel free to ask any questions or use this code for your own projects.
Matthias