Project: Panasonic 3D Lidar Challenge - olinrobotics/gravl GitHub Wiki
Overview
Fall 2019
Spring 2020
Tasks
Priority: HIGH
- Fix OPC & Safety System
- Verify & Re-tune Teleoperative Capability
- Train ML Algorithms on new data collected from Waymo-OD
- Update Panasonic Lidar transform
- Update person-follow node to follow point in 2D space
- Write Panasonic Lidar node to output person location as point in 2D space
- Test batching algorithm on simulated data
- Using 3D Lidar to detect road height
- Collect training data from virtual 3D Lidar
Priority: MEDIUM
- Write script to generate training data from Panasonic Lidar
- Collect training data of cones to augment Waymo Dataset
- Retrain ML Algorithms on additional cone data
Priority: LOW
Extensions
- Combine with CV algorithm to resolve multi-person confusion
- Set type of object to follow, allow tractor to follow vehicles