Marius Memmel
@memmelma.bsky.social
Robotics PhD student @UW and Intern @Bosch_AI Pittsburgh 🐺 previously @DHBW @EPFL, @TUDarmstadt 🇩🇪
Interested in retrieving (sub-)trajectories from enormous datasets like DROID? 🦣
Check out our retrieval code using DTW and vision foundation model features!
Code: github.com/WEIRDLabUW/S...
Website: weirdlabuw.github.io/strap
Check out our retrieval code using DTW and vision foundation model features!
Code: github.com/WEIRDLabUW/S...
Website: weirdlabuw.github.io/strap
STRAP retrieves semantically meaningful trajectories on large datasets like DROID making it a powerful general-purpose algorithm for robot trajectory retrieval!
🧵 5/6
🧵 5/6
February 12, 2025 at 8:38 AM
Interested in retrieving (sub-)trajectories from enormous datasets like DROID? 🦣
Check out our retrieval code using DTW and vision foundation model features!
Code: github.com/WEIRDLabUW/S...
Website: weirdlabuw.github.io/strap
Check out our retrieval code using DTW and vision foundation model features!
Code: github.com/WEIRDLabUW/S...
Website: weirdlabuw.github.io/strap
Reposted by Marius Memmel
Happy holidays from UW Robotics!
December 24, 2024 at 6:15 PM
Happy holidays from UW Robotics!
Have some offline data lying around? Use it to robustify few-shot imitation learning! 🤖
STRAP 🎒 is a retrieval-based method that leverages semantic sub-trajectories in offline datasets to augment the training data.
🧵 1/6
STRAP 🎒 is a retrieval-based method that leverages semantic sub-trajectories in offline datasets to augment the training data.
🧵 1/6
December 20, 2024 at 11:28 PM
Have some offline data lying around? Use it to robustify few-shot imitation learning! 🤖
STRAP 🎒 is a retrieval-based method that leverages semantic sub-trajectories in offline datasets to augment the training data.
🧵 1/6
STRAP 🎒 is a retrieval-based method that leverages semantic sub-trajectories in offline datasets to augment the training data.
🧵 1/6
Reposted by Marius Memmel
I'm excited about scaling up robot learning! We’ve been scaling up data gen with RL in realistic sims generated from crowdsourced videos. Enables data collection far more cheaply than real world teleop. Importantly, data becomes *cheaper* with more environments and transfers to real robots! 🧵 (1/N)
December 5, 2024 at 2:13 AM
I'm excited about scaling up robot learning! We’ve been scaling up data gen with RL in realistic sims generated from crowdsourced videos. Enables data collection far more cheaply than real world teleop. Importantly, data becomes *cheaper* with more environments and transfers to real robots! 🧵 (1/N)