Integrated Data-driven Inference and Planning-based Human Motion Prediction for Safe Human-Robot Interaction
We develop the algorithm for a autonomous vehicle to safely and actively interact with an uncertain human-driven vehicle.
The algorithm combines:
- A hierarchical prediction strategy that integrates data-driven human internal state inference with planning-based human motion prediction
- An Active motion planning algorithm for the autonomous vehicle to ensure safety against uncertain human motions
Y. Nam and C. Kwon, “Integrated Data-driven Inference and Planning-based Human Motion Prediction for Safe Human-Robot Interaction”, ICRA 2024: International Conference on Robotics and Automation, Yokohama, Japan, May 2024
- Generate human driving data for training
-> python ./gen_human_data.py - Train rationality inference module
-> python ./train_inference/train_beta.py - Train driving style inference module
-> python ./train_inference/train_psi.py
- Run main code
-> python ./main.py
** After training the inference module, 'model_id' for each inference module should be changed accordingly.