Tara Sadjadpour

CS PhD @berkeley_ai. Prev: @ToyotaResearch, @nvidia, MS @StanfordAILab, BS @UCLA

[Tara Sadjadpour @TaraSadjadpour]

One human demonstration. Any multi-fingered hand. Zero-shot sim-to-real visuomotor policy.

https://t.co/fqkm4Fc1qv

Collaborators: @he_siming @ckwolfeofficial @HaozhiQ @LeaMue27 Shankar Sastry, Claire Tomlin, @JitendraMalikCV

[Sharpa @SharpaRobotics]

We’re glad to see @TaraSadjadpour and the Morphometric Imitation team advancing research in morphology adaptation, one-shot learning, and sim-to-real transfer with our 22-DoF dexterous hands.

Sharpa provides simulation assets with faithful kinematics and dynamics, while its human-like morphology enables a broad range of dexterous motions.

Interested in exploring it yourself? Get in touch or download the simulation assets from our website: https://t.co/lmHm70mPXh

[Siming He @he_siming]

🎉 Workshop on Sim-to-Real-to-Field: The Science of Transfer is coming to #CoRL2026 in Austin!

Our CFP is now open — with a $1,000 Best Paper Award generously sponsored by @Tensr.

🔗 https://t.co/ERunSUrmtv

[Tara Sadjadpour @TaraSadjadpour]

Shoutout to prior work that helped this project:
@omidtaherii et al for the GRAB dataset (https://t.co/YlBSvCXq4U)
@Stone_Tao et al for ManiSkill (https://t.co/9mvKjJRrMw)
@brenthyi for jaxls (https://t.co/ao9sAwcCpm) and MANO JAX implementation (https://t.co/37u6nserip)

[Tara Sadjadpour @TaraSadjadpour]

One human demonstration. Any multi-fingered hand. Zero-shot sim-to-real visuomotor policy.

https://t.co/fqkm4Fc1qv

Collaborators: @he_siming @ckwolfeofficial @HaozhiQ @LeaMue27 Shankar Sastry, Claire Tomlin, @JitendraMalikCV

[Neerja Thakkar @neerjathakkar]

What’s the right representation for a world model? 3D, pixels, or something else?

Excited to release our new paper “Forecasting Motion in the Wild” where we propose point tracks as tokens for generating complex non-rigid motion and behavior

From @GoogleDeepmind @Berkeley_AI @TTIC_Connect

[Tara Sadjadpour @TaraSadjadpour]

Check out PAR @CVPR!

We apply PAR to 3 case studies: social interaction, driving, & HOI. These results can be useful in downstream applications, like robotics & autonomous driving.

The code is easy to try with other datasets as a baseline to build from: https://t.co/EgRCTLPeGO

[Neerja Thakkar @neerjathakkar]

Can we systematically generalize AR "word models" into "world models”? Our CVPR 2025 paper introduces a unified, general framework designed to model real-world, multi-agent interactions by disentangling task-specific modeling from behavior prediction.