VIGOR-MPC: Vision-Guided Objective Regulation MPC
VIGOR-MPC: Vision-Guided Objective Regulation MPC
This paper presents VIGOR-MPC, a modular student-teacher framework for vision-guided control in autonomous navigation. The system distills a long-horizon MPC into a computationally lightweight short-horizon MPC, using online vision-based cost parameter estimation. Evaluated in simulation and on hardware, VIGOR-MPC generalizes to unseen tracks and achieves lap times close to the teacher model while reducing computation by over 90%.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
- Environment
- SimulationOpen source S4
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
ves only the noisy, occluded context produced by a perception module and learns, via a multi-level recurrent mapping, to infer cost parameters that aim to recover the teacher's behavior despite missing information. We evaluate VIGOR-MPC on vision-based autonomous racing in simulation and on hardware. Using only onboard perception of the track, VIGOR-MPC generalizes to tracks unseen during training and approaches the
Open source S4
Source：arXiv Robotics — research abstracts · arxiv.org