UniBwTAS/CollisionPro
Towards explainable value functions in reinforcement learning. A framework for collision probability distribution estimation via deep temporal difference learning.
This project helps engineers and researchers in robotics or autonomous systems understand and predict collision risks. It takes in sensor data or simulated environment states and outputs a clear probability distribution of potential collisions, allowing users to make more informed decisions about system safety and behavior. It's designed for those who need transparent risk assessments, not just a pass/fail.
No commits in the last 6 months.
Use this if you need to go beyond simple collision detection and require an interpretable, probabilistic understanding of collision risks in dynamic environments, especially for systems using reinforcement learning.
Not ideal if you are looking for a simple, real-time collision avoidance system that doesn't require deep analysis of probability distributions.
Stars
14
Forks
2
Language
Python
License
BSD-3-Clause
Category
Last pushed
May 05, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/UniBwTAS/CollisionPro"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
microsoft/AirSim
Open source simulator for autonomous vehicles built on Unreal Engine / Unity, from Microsoft AI...
lgsvl/simulator
A ROS/ROS2 Multi-robot Simulator for Autonomous Vehicles
microsoft/AirSim-NeurIPS2019-Drone-Racing
Drone Racing @ NeurIPS 2019, built on Microsoft AirSim
DeepTecher/AutonomousVehiclePaper
无人驾驶相关论文速递
learn-to-race/l2r
Open-source reinforcement learning environment for autonomous racing — featured as a conference...