Variational Autoencoders for exteroceptive perception in reinforcement learning-based collision avoidance

arXiv:2404.00623v1 Announce Type: new
Abstract: Modern control systems are increasingly turning to machine learning algorithms to augment their performance and adaptability. Within this context, Deep Reinforcement Learning (DRL) has emerged as a promising control framework, particularly in the domain of marine transportation. Its potential for autonomous marine applications lies in its ability to seamlessly combine path-following and collision avoidance with an arbitrary number of obstacles. However, current DRL algorithms require disproportionally large computational resources to find near-optimal policies compared to …

abstract adaptability algorithms applications arxiv autoencoders autonomous collision context control control systems cs.lg cs.ro domain framework lies machine machine learning machine learning algorithms marine modern perception performance reinforcement reinforcement learning systems transportation type variational autoencoders