A group project implementing autonomous parking in Unity Engine.
Key Points
iPark: Intelligent Parking is a machine learning project that utilizes Unity ML-Agents, Reinforcement Learning (RL), and the Unity game engine to create an intelligent system capable of parking a car in various dynamic scenarios.
It can work in 2 modes: Training and Evaluation. When the project is run in Unity Engine alongside Unity ML-Agents, it can be used to train a model for the task of car parking. When the executable or exported file is run, it works in model evaluation mode.
In evaluation, the agent autonomously parks the car in a random empty slot avoiding any collisions.
In training, the agent learns this behavior through Reinforcement Learning techniques. Precisely, it uses Proximal Policy Optimisation (PPO) alogorithm provided by the Unity ML-Agents.
It is a group project led by Kushagra and is the foundation of a
Research Paper.
The project report can be found
here.
More details on the components and working can be provided upon request.