avisingh599/reward-learning-rl

[RSS 2019] End-to-End Robotic Reinforcement Learning without Reward Engineering

48
/ 100
Emerging

This project helps robotics engineers train robots to perform complex manipulation tasks like draping, pushing, or opening doors. It takes in visual observations (pixels) from the robot's camera and a small number of example images showing the desired goal state. The output is a trained robot policy that can autonomously complete the specified task without needing manual reward programming.

375 stars. No commits in the last 6 months.

Use this if you need to teach a robot new manipulation skills from visual input and want to avoid the tedious and complex process of hand-crafting reward functions for each task.

Not ideal if your robotic task is simple and well-defined enough for traditional programming, or if you don't have visual examples of the desired goal state.

robot-manipulation robot-training automation robotics-engineering machine-vision
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 22 / 25

How are scores calculated?

Stars

375

Forks

69

Language

Python

License

Last pushed

Nov 22, 2022

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/avisingh599/reward-learning-rl"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.