Simon Guist retweeted
🤖 We’re hiring 2 PhD students at our brand-new Institute for Machine Intelligence, JKU Linz!
Our mission: push the boundaries of robotics by building the next generation of more capable robots, inventing new learning methods that bring them to life, and everything in between.
🧠 Robot learning → tinyurl.com/ypbs7v9w
🦾 Robot design for robot learning → tinyurl.com/2n99thx4
Join as one of our very first members and help shape the institute from day one. Deadline is in a week!
RTs appreciated 🙏
Simon Guist retweeted
First sim-to-real transfer for muscle-actuated robots 🦾
Muscle-actuated robots are powerful but hard to model. GeAN addresses this longstanding challenge by learning muscle and tendon dynamics, enabling sim-to-real transfer.
arxiv.org/abs/2604.09487
youtu.be/EOgdJxghYaw
🧵👇
Simon Guist retweeted
Excited to welcome Dieter Buchler @dtrbchlr to the University of Alberta! We are bringing robotics and reinforcement learning together for an awesome, interactive, and experiential AI.
A bit late but still want to congratulate @RichardSSutton for his achievements in RL and AI & his huge positive influence on so many researchers!
Guess it’s a good moment to announce that I joined @UAlberta 😎 Feeling honored to work alongside Rich and my other colleagues!
Simon Guist retweeted
Replying to @simonguist
I’m very proud that we can finally present this project at #RSS2024. We iterated on the hardware over the past 3 years to deliver a truly robust *and* fast open-source tendon-driven arm. Robot bodies play an essential role in learning control!
We presented our work on a novel open-source (mostly) 3D-printable tendon-driven robot arm at #RSS2024 today. Our design enables safety through reduced inertia and passive compliance, while addressing challenges regarding friction and robustness
youtube.com/watch?v=5vH2_TAa…
Experiments include a high-speed table tennis task (ball speeds up to 20 m/s). Open-source hardware, software, and a motion dataset: webdav.tuebingen.mpg.de/pamy…
1
2
117
This was truly a collaborative effort, many thanks to my brilliant colleagues @JanS1854, Hao Ma, @clthegoat, Vincent Berenz, Julian Martus, Heiko Ott, Felix Grüninger, Michael Muehlebach, Jonathan Fiene, @bschoelkopf and @dtrbchlr
2
97
Simon Guist retweeted
Gradient subspace optimization unlocked for RL 🔒➡️🔓
Used only for supervised learning so far, our #ICLR2024 paper illustrates that policy gradients evolve in a small, slowly-changing subspace, opening up many opportunities for more efficient RL.
arxiv.org/abs/2401.06604
Robots are limited by reality, but everything is possible in virtual environments.
Our #RSS2023 paper presents Hindsight States (HiS), a method that leverages imbalances in complexity between parts of the task for efficient learning. (1/4)
We show how this approach can accelerate learning both on its own and when combined with Hindsight Experience Replay (HER). Check out how we learned to play table tennis on a real muscular robot with multiple virtual balls. (3/4)
1
2
245
Title: ‘Hindsight States (HiS): Blending Sim and Real Task Elements for Efficient Reinforcement Learning'
arxiv.org/abs/2303.02234
Work done at @MPI_IS with my brilliant colleagues Jan Schneider, @alexdttrch, Vincent Berenz, @bschoelkopf and @dtrbchlr (4/4)




