@KyleMorgenstein Check out the OGMP line of work. We do guided exploration!
Oracle Guided Multi-mode Policies https://t.co/8qpCG7xYyj
Preferenced OGMP (for loco-manipulation)
https://t.co/8CgBYsu94J
Our work "Preferenced Oracle Guided Multi-mode Policies for Bipedal Loco-Manipulation" is accepted at IROS 2025 & Whole-body Control and Bimanual Manipulation workshop at RSS!
Catch us at the poster session - Wed, 10:40 am @USC#RSS2025
Isaac Lab codebase for Oracle Guided Multi-mode Policies with Mode-Preference is now open!
Check it out here: https://t.co/r5ygKhqG5G
Try our policies with Unitree G1, H1, or Berkeley Humanoid!
Check out the paper and feel free to reach out! Incredibly grateful to my collaborators @lok_s_h and @NikhilSoban353 at the Dynamics Robotics and Control Laboratory (DRCL) @USC. Special thanks to Prof. Quan Nguyen @qnguyen_usc for his valuable inputs throughout the project! (6/n)
Policies may get stuck in local optima by choosing an undesirable sequence of modes. To mitigate this, we introduce a task-agnostic mode-switching preference reward using mode ranks specified by the user. (5/n)
@animesh_garg Interesting line of work @imgeorgiev@krishpopdesu, excited to see where it goes!
Do you have any comparisons on contact-rich manipulation tasks? Truncating the gradients at contact may make such tasks difficult as all the interesting & difficult stuff happens at contact.
if anyone is looking for QEC or atomic physics researcher roles and want to join the core team of an exciting QC startup (I'm not the founder xD), feel free to DM me.
#qec#quantum#QuantumComputing
The worlds' first free medical college
1/9 Did you know that India produces the largest number of #Doctors in the world?
There are more than 550 allopathic #Medical colleges and 90,000 #MBBS graduates every year.