Michiel van de Panne | SCA 2020: ALLSTEPS Curriculum-driven Learning of Stepping Stone Skills (3 minute overview) @m.vandepanne | Uploaded September 2020 | Updated October 2024, 2 hours ago.
A short overview of the paper "ALLSTEPS: Curriculum-driven Learning of Stepping Stone Skills", ACM/EG Symposium on Computer Animation (SCA 2020).
Project page:
cs.ubc.ca/~van/papers/2020-allsteps/index.html
Special thanks to STORROR Pro Parkour Team (storror.com), Jesse Chui (songsforcinema.com), and the Proclaimers (excerpts from "I'm Gonna Be (500 miles) the.proclaimers.co.uk/).
A short overview of the paper "ALLSTEPS: Curriculum-driven Learning of Stepping Stone Skills", ACM/EG Symposium on Computer Animation (SCA 2020).
Project page:
cs.ubc.ca/~van/papers/2020-allsteps/index.html
Special thanks to STORROR Pro Parkour Team (storror.com), Jesse Chui (songsforcinema.com), and the Proclaimers (excerpts from "I'm Gonna Be (500 miles) the.proclaimers.co.uk/).
![[SIGGRAPH 2020] Character Controllers using Motion VAEs](https://i.ytimg.com/vi/Zm3G9oqmQ4Y/hqdefault.jpg "[SIGGRAPH 2020] Character Controllers using Motion VAEs
Video for the related SIGGRAPH 2020 paper
A fundamental problem in computer animation is that of realizing purposeful and realistic human movement given a sufficiently-rich set of motion capture clips. We learn data-driven generative models of human movement using autoregressive conditional variational autoencoders, or Motion VAEs. The latent variables of the learned autoencoder define the action space for the movement and thereby govern its evolution over time. Planning or control algorithms can then use this action space to generate desired motions. In particular, we use deep reinforcement learning to learn controllers that achieve goal-directed movements. We demonstrate the effectiveness of the approach on multiple tasks. We further evaluate system-design choices and describe the current limitations of Motion VAEs.")
