Brent Yi
@brenthyi
Followers
1K
Following
1K
Media
9
Statuses
98
RT @marionlepert: Introducing Masquerade 🎭: We edit in-the-wild videos to look like robot demos, and find that co-training policies with th….
0
35
0
RT @QianqianWang5: 📢Thrilled to share that I'll be joining Harvard and the Kempner Institute as an Assistant Professor starting Fall 2026!….
0
40
0
RT @qiayuanliao: Want to achieve extreme performance in motion tracking—and go beyond it? Our preprint tech report is now online, with open….
0
239
0
RT @Almorgand: "Cameras as Relative Positional Encoding".TLDR: comparison for conditioning transformers on cameras: token-level raymap, att….
0
52
0
RT @redstone_hong: 🤖 Initial code release is up for VideoMimic Real2Sim!. VideoMimic is a real-to-sim-to-real pipe….
github.com
Visual Imitation Enables Contextual Humanoid Control. arXiV, 2025. - hongsukchoi/VideoMimic
0
35
0
Such an impressive code release!!.
@brenthyi @arthurallshire VideoMimic makes extensive use of Viser for rich 3D visualizations in the real-to-sim pipeline. Check it out:
2
14
178
RT @zhenjun_zhao: Viser: Imperative, Web-based 3D Visualization in Python. @brenthyi, @ChungMinKim, @justkerrding, Gina Wu, Rebecca Feng, A….
0
23
0
Thanks also @arthurallshire for contributing VideoMimic results to the demo reel!.
our new system trains humanoid robots using data from cell phone videos, enabling skills such as climbing stairs and sitting on chairs in a single policy. (w/ @redstone_hong @junyi42 @davidrmcall)
1
2
18
Docs: Paper: This has been a few years in the making with a fantastic group of people 💕. cc @ChungMinKim, @jonaskulhanek, @justkerrding, @redstone_hong, @YiMaTweets, @akanazawa, and many many others
1
5
43
July has been a big month for Viser!.- Released v1.0.0😊.- We did some writing. Some demos👇
14
101
761
And initial code release is up! Really enjoyed working with @kevin_zakka et al's MuJoCo Playground for this project⭐.
github.com
Implementation of Flow Policy Optimization (FPO) . Contribute to akanazawa/fpo development by creating an account on GitHub.
Excited to share Flow Matching Policy Gradients: expressive RL policies trained from rewards using flow matching. It’s an easy, drop-in replacement for Gaussian PPO on control tasks.
3
22
223
RT @YouJiacheng: this guy is really good at makiong cool AND useful blogs (and research!).DDM is also very nice.
0
3
0
RT @ruilong_li: GTA deserves way more attention! 🔥.It’s not just for 3D vision—rotating Q, K *and more importantly V and O* with RoPE-style….
0
12
0
RT @HaozhiQ: Our code has been released: Check out how we simulate the magnetic tactile skin and transfer it to th….
github.com
Contribute to jessicayin/tactile_skin_model development by creating an account on GitHub.
0
14
0
RT @YiMaTweets: One step further towards better pose-controlled 3D generation. .
0
3
0