Zhu-H-Y / RealScienceComments

0 stars 0 forks source link

Diffusion Policy: Visuomotor Policy Learning via Action Diffusion #12

Open utterances-bot opened 8 months ago

utterances-bot commented 8 months ago

Diffusion Policy: Visuomotor Policy Learning via Action Diffusion

Robotics: Science and Systems (RSS) 2023

https://real-science.vercel.app/Diffusion%20Policy:%20Visuomotor%20Policy%20Learning%20via%20Action%20Diffusion

HaoyiZhu commented 8 months ago

Rate: 4

Diffusion policy might be currently the most popular BC algorithm. BC from teleop demonstrations has many challenges such as non-markovian, non-stationary, sub-optimal, etc. The generative modeling can naturally adress these problems and perform very well.