Yuemin Mao

PhD student @CMU_Robotics

[Yuemin Mao @YueminMao]

VibeAct lets your dexterous robot hand "hear" and react to contact events!

With real-world acoustic data and digital-clone labeling, VibeAct learns a physics-grounded tactile representation that bridges sensing and simulation, enabling reactive policies trained in sim to transfer directly to the real robot.

VibeAct is an exciting project co-led by @UksangYoo and me! See the project website (https://t.co/TN2R1BeX3a) and Uksang's thread for more details.

[Uksang Yoo @UksangYoo]

How can we get robot hands to “hear” slip and contact through microphones, and react to them?

We’re excited to share VibeAct, an approach that uses piezoelectric microphones embedded in robot fingertips to estimate contact and slip, then learns reactive policies from this tactile feedback!

https://t.co/7jMLZNpYzp

[Krishna Suresh @_krishnasuresh]

Check out our recent work accepted to #RSS2026! We enable a robot to learn a flying knot from a single human demonstration and less than 10 trials using Task-Level Iterative Learning Control:
https://t.co/8ZAi3PIbzY

[Binghao Huang @binghao_huang]

1/ 🤲 LeRobot has made low-cost robot learning widely accessible — but most policies are still blind to contact.

Today we release LeFlexiTac: a tactile extension for the LeRobot platform using FlexiTac sensors. Make tactile robot learning as easy as possible.

Project page: https://t.co/6PY8oTmAjU
Code/docs: https://t.co/11HW0Zwrtb

[Yaru Niu @yaru_niu]

A touch-aware humanoid manipulation policy that cleans the lab for you🧹🧪

Introducing Humanoid Touch Dream: a real-world system for dexterous, contact-rich humanoid loco-manipulation.

Our key idea is simple: the policy predicts future hand forces and tactile latents alongside actions, within a single-stage training framework.

https://t.co/Pt5pXA65wm

1/7

[Yuemin Mao @YueminMao]

That’s a really interesting direction to explore. In this work, we focused on objects that do not generate sound themselves, but we did study whether the model is capturing the right signals for slip. Specifically, we compared performance with the robot off, on but stationary, and moving, where the motors introduce different levels of noise. We found that including a small dataset with different noise conditions in the finetuning step improves model robustness.

I would expect that finetuning with objects that produce sound could similarly help the model handle these cases. This would be a valuable direction for future work.

[Yuemin Mao @YueminMao]

Let your robots hear slips with A-SLIP! 🤖🎧

How can a robot detect in-hand slip and estimate its direction and magnitude without cameras or fragile tactile skins?

A-SLIP uses piezoelectric microphones embedded in grippers to hear it.

https://t.co/VZWCZ7XaPe

🧵1/7

[Yuemin Mao @YueminMao]

Yeah that’s a great point! Structure-borne vibrations carry rich information that extends beyond slip estimation. That said, training a model to identify cracking sound or classify different contact events and materials is definitely possible, and these learned representations are useful for downstream tasks like reactor control.

[Yuemin Mao @YueminMao]

A-SLIP shows that passive structure-borne acoustics, combined with spatially distributed microphones and learning-based fusion, is a practical path to continuous slip estimation without the cost, bulk, or fragility of vision-based tactile sensors.

A-SLIP is co-led by @UksangYoo, advised by @jeanoh and @jeff_ichnowski at @CMU_Robotics, and supported by @Samsung_RA. Big thanks!👏

Website: https://t.co/IUK8YOsMha
arXiv: https://t.co/vcWP7iTwzC
(Code + Dataset coming soon)

🧵7/7

[Yuemin Mao @YueminMao]

We test A-SLIP in two closed-loop tasks:

1) In Slip-Stop, the robot pushes objects into a wall and halts on slip detection. A-SLIP gets 50/50 successes vs 31/50 for the baseline, with 45% lower stopping error.
2) In Slip-Track, the robot follows externally induced slip to maintain grasp pose, achieving 50% lower trajectory RMSE.

🧵6/7

[Uksang Yoo @UksangYoo]

Excited to share SoftAct, a framework for retargeting human manipulation demos to soft robot hands using explicit contact force reasoning! How do you transfer human skill to a hand that looks and moves nothing like yours🐙🖐️? It turns out VR environments can let us capture privileged force interaction demonstrations to help. 🧵1/7

[Adam Hung @Adamjhung]

Learning from human videos often requires restrictive, carefully choreographed human motions.

We propose ✨3PoinTr✨: a scalable way to pretrain from casual human videos. It bridges the embodiment gap by learning 3D scene evolution, enabling learning from natural human motions.

[Hongyi Chen @chen_hongyi_]

How far can we push dexterous robot manipulation with human video-only supervision and minimal assumptions?
🚫 No teleop. 🚫 No wearables. 🚫 No external sensors. 🚫 No robot demos.
Introducing VIDEOMANIP: 🎥 Just monocular RGB, 🌍 in-the-wild human video → dexterous robot manipulation 🤚[1/6]

[Yu Chen @YuChen97026150]

🤖🦾✍️Why is robot grasping hard? We usually blame contacts, kinematics, geometries, perception, and so on. But what if the object is just being spiteful?

🔥We propose a game-theoretic grasp synthesis method as a two-player game between the robot and an adversarial (spiteful) object.

💡In this formulation we achieve SOTA grasp success rates, without training data - just using optimization tools (Augmented Lagrangian + Iterative Best Response).

📑Arxiv: https://t.co/mH2ocISuHy
🌐Website: https://t.co/HKayRv9yRD

🧵[1/7]

[Yuemin Mao @YueminMao]

@IlirAliu_ Thanks so much for sharing this work!!!

[Chengyang Zhao @alecyzh]

🤖What if a robot could understand hair dynamics well enough to style your hair, just like your favorite barber💈?

🔥Excited to announce DYMO-Hair, a model-based robot hair styling system powered by a generalizable 3D hair dynamics model.

🚀A new step toward robots that can understand and manipulate more complex deformable materials.

New dynamics learning paradigm. Fully synthetic data from a novel lightweight simulator. Zero-shot sim2real transfer.
👉Check it out at: https://t.co/Kx2gyWBAAD
More details below: [1/N]

[Matthew Mason @mastertoadster]

Time for blog post number eight -- "Vacuum grippers versus robot hands." Here's the question: when humanoids can do everything a human can, will they take over the warehouse? No! Vacuum is awesome! Vacuum grippers are the wheels of manipulation!

It is only an five-minute read. Hope you like it.

https://t.co/feOTrKdKcO