/AI1h ago

TraceGen And SIMPACT Improve Robot Learning With 3D Traces And Simulation

27161.8K
#321Jia-Bin Huang
Original post
Jia-Bin Huang@jbhuang0604#321inAI

Robot learning is bottlenecked by data: adapting to new tasks, robots, and environments often requires many demonstrations.

Meet TraceGen! A motion world model that learns in a compact 3D trace space, abstracting away complex appearance.

This enables learning from diverse videos.

Meet @JayLEE_0301 at #605!

Jia-Bin Huang@jbhuang0604

VLMs are great for their semantic reasoning capabilities.

BUT, they don't have a grounded understanding of physical dynamics.

So, we equip VLMs with physical reasoning through simulation-in-the-loop world modeling.

Come talk with Haowen about SIMPACT (#611): https://simpact-bot.github.io/

6:59 AM · Jun 6, 2026 · 830 Views
Sentiment
Sentiment building, check back later.
Cluster Engagement
Posts from X
Most Activity
Most Activity
VIEWS932BOOKMARKS4LIKES6RETWEETS1REPLIES1
Jia-Bin Huang@jbhuang0604

VLMs are great for their semantic reasoning capabilities.

BUT, they don't have a grounded understanding of physical dynamics.

So, we equip VLMs with physical reasoning through simulation-in-the-loop world modeling.

Come talk with Haowen about SIMPACT (#611): https://simpact-bot.github.io/

Jia-Bin Huang@jbhuang0604

Hi friends at #CVPR2026! 👋

Please come check out our research today!

@YaoChihLee will showcase a super-fun video motion editing work - Edit-by-Track (#378)!

1hViews 932Likes 6Bookmarks 4
Jia-Bin Huang@jbhuang0604

Customizing multiple visual concepts remains challenging.

We introduce UniVerse, a segmentation-free framework that learns to disentangle, localize, and recombine individual concepts within diffusion transformers.

Come check out the cool results by Quynh (#53). https://universe-personalization.github.io/

1hViews 108Likes 1Bookmarks 1
Jia-Bin Huang@jbhuang0604

I will be around all poster sessions!

Please come say HI!

1hViews 100Likes 1