I-Chun (Arthur) Liu

CS (Robotics & AI) PhD Candidate @USC | Prev. Intern @ Amazon Robotics.

[I-Chun (Arthur) Liu @arthur801031]

Excited to share that I'll be presenting our work on data augmentation for eye-in-hand bimanual imitation learning at CoRL (@corl_conf) next week! Thanks to the team @jasoon_chen, @gauravsukhatme, @daniel_t_seita at @CSatUSC.

[USC RESL @uscresl]

Learning-based systems still require large amounts of data for effective generalization. Our CoRL 2025 paper, D-CODA, expands data without simulators or environment interactions by generating novel view-consistent wrist-camera images and action labels for bimanual robots.
🧵👇

[USC RESL @uscresl]

Learning-based systems still require large amounts of data for effective generalization. Our CoRL 2025 paper, D-CODA, expands data without simulators or environment interactions by generating novel view-consistent wrist-camera images and action labels for bimanual robots.
🧵👇

[I-Chun (Arthur) Liu @arthur801031]

I’ll be presenting VoxAct-B at @corl_conf this afternoon during Poster Session 4, from 3:30 PM to 5:30 PM, at poster #18. Come check out our poster!

🧵: https://t.co/aHCusfGP4s
Website: https://t.co/ytu0XolH3a

[USC RESL @uscresl]

Tasks requiring two-hand coordination and fine-grained manipulation remain challenging for current robotic systems. Our CoRL 2024 paper proposes a sample-efficient, language-conditioned, voxel-based method that utilizes Vision Language Models to address these challenges.

🧵👇

[I-Chun (Arthur) Liu @arthur801031]

Excited to share that I will be presenting our work on bimanual manipulation at CoRL (@corl_conf) next week!

Thanks to the team @SichengHe12345, @daniel_t_seita, @gauravsukhatme at @CSatUSC.

[USC RESL @uscresl]

Tasks requiring two-hand coordination and fine-grained manipulation remain challenging for current robotic systems. Our CoRL 2024 paper proposes a sample-efficient, language-conditioned, voxel-based method that utilizes Vision Language Models to address these challenges.

🧵👇

[USC RESL @uscresl]

Tasks requiring two-hand coordination and fine-grained manipulation remain challenging for current robotic systems. Our CoRL 2024 paper proposes a sample-efficient, language-conditioned, voxel-based method that utilizes Vision Language Models to address these challenges.

🧵👇

[USC RESL @uscresl]

Excited to share MAIL, a new framework that enables robots to learn from demos by robots with different morphologies! 🎉
Website: https://t.co/EIsFFVJRsp
Paper: https://t.co/n499yOaiT3
Presentation: next week @corl_conf

A thread🧵, 1/9

[USC Thomas Lord Department of Computer Science @CSatUSC]

It's official! 🎉 We are the Thomas Lord Department of Computer Science! 👏

Congratulations Professor @gauravsukhatme, who was named the inaugural Donald M. Alstadt Chair in Advanced Computing. 

Thank you to all who made this such a memorable event. More to come! @USCViterbi

[USC RESL @uscresl]

Now accepted to #CoRL2023! With morphological adaptation in #ImitationFromObservation, we teach one-armed robots tasks that you can easily demonstrate bimanually. Watch this space for more updates!
@GautamSalhotra @arthur801031 @gauravsukhatme @corl_conf

[USC RESL @uscresl]

Our paper on #DeformableManipulation (DMfD) is accepted to #IEEERAL & #IROS22! We learn manipulation of 1D & 2D deformables (rope, cloth) using images, with 6% sim2real gap
https://t.co/pswoOXO42O
https://t.co/2eKfX2mL0r
@GautamSalhotra @arthur801031 @marcus_kuhne @gauravsukhatme

[Gautam Salhotra @GautamSalhotra]

Our paper on learning deformable object manipulation using expert demonstrations has been accepted to Robotics & Automation Letters (RA-L) by @ieeeras !
With coauthors @arthur801031, @marcus_kuhne, @gauravsukhatme at @uscresl. Manuscript and other details coming soon!