policy

RoboVision

Gautam-R-Patil · PyTorch

or hover any field below to flag it

Overview

Name
RoboVision
Author
Gautam-R-Patil
Framework
PyTorch
License
unknown
Skill type
locomotion
Evidence level
untested
Task description
Developed a physics-grounded locomotion policy for a Unitree quadruped in the Genesis environment using Deep RL (PPO). Trained the agent to autonomously walk, run, and rotate on its yaw axis via optimized reward functions and continuous proprioceptive feedback.

Spaces

Action space
other · 0-dim · 0Hz
Observation space
  • type: other

Links

HuggingFace repo
null
Paper (arXiv)
null

Compatible robots

0+20 mentioned but not in catalog yet

No robots list RoboVision as compatible yet. Know of one? Flag it above.

Compatible environments

0

No environments list RoboVision yet.

Datasets that reference this policy

0

No datasets reference RoboVision yet.