Research intern @ Amazon FAR | PhD student @ Cornell | ๐๐๏ธ๐๐ค
SH โ๏ธ NYC โ๏ธ SF
Joined March 2022
- Tweets118
- Following1.1K
- Followers1.1K
- Likes355
Pinned Tweet
Introducing DreamFAR, a world model trained from scratch using only robot data, without internet pretraining. It supports:
๐ฎ Real-time teleoperation
โฑ๏ธ Minutes of long-horizon rollout
๐งฉ Generalization across 1,000+ tasks
yimingdou.com/DreamFAR
๐งต (1/8)
Work done during internship at Amazon FAR w/ @BotaoUMD @antoniloq @andrewhowens @JitendraMalikCV @SOTA_kke @ruoshi_liu
Special shout out to my wonderful mentor @ruoshi_liu for so much inspiration on the high-level idea and detailed discussions on the implementation!
Yiming Dou retweeted
Human videos bring scale and motion priors to robot pretraining. But video mainly captures human states, leaving intended actions implicit. Applied forces help fill that gap. Introducing FACT: learning dexterous manipulation from human video and force. ๐งต
Yiming Dou retweeted
Thrilled to share CARD: Contact-Anchored Retargeting for Dexterous manipulation. So far, collecting robot data, teleop, UMI, kinesthetic teaching, has run into the same two problems: 1) the mapping is largely kinematics only, and 2) the human has to act through the robot and hope the interaction works out. CARD instead treats contact and forces as the interface, from recording to replay, which is what lets it handle dexterous tasks that motion-only methods struggle with.
๐งต below 1/
Can you believe these are open-loop rollouts?
Excited to introduce CARD ๐ โ Contact-Anchored Retargeting for Dexterous Manipulation.
Almost every method today is limited to kinematics alone as the ๐ข๐ง๐ญ๐๐ซ๐๐๐๐ between the human and the robot. With CARD, we track real world contact and forces alongside kinematics to perform unprecedented dexterous tasks through tactile reactive replay!
CARD is scalable:
๐น Multiple input sources
๐ค Multiple robot hands
โก needs no object models, no simulator, 10 Hz
Yiming Dou retweeted
Can you believe these are open-loop rollouts?
Excited to introduce CARD ๐ โ Contact-Anchored Retargeting for Dexterous Manipulation.
Almost every method today is limited to kinematics alone as the ๐ข๐ง๐ญ๐๐ซ๐๐๐๐ between the human and the robot. With CARD, we track real world contact and forces alongside kinematics to perform unprecedented dexterous tasks through tactile reactive replay!
CARD is scalable:
๐น Multiple input sources
๐ค Multiple robot hands
โก needs no object models, no simulator, 10 Hz
Yiming Dou retweeted
Collecting data for dexterous hands is hard. Teleop needs a full setup. Bringing UMI-like devices to 20-dof hands is challenging.
Introducing DexPort ๐๏ธ: a hand-held interface w/ a real robot hand. No embodiment gap, recorded joint state AND actions, easy to use.
Yiming Dou retweeted
Your robot's shape shouldn't limit its sense of touch.
Meet OpenTac: an open ecosystem for custom tactile skins that sense normal + shear forces.
Very excited about this collaborative project! Canโt wait to see what the community builds with it. ๐งต๐
opentac.github.io
Yiming Dou retweeted
Robots canโt always see the physics they act under: hidden mass, friction, kinematic constraints. A scalar reward only says "failed." The outcome of an action says much more.
Introducing SCOUT: test-time policy adaptation from action-outcome feedback. Sim-trained, adapts in the real world in a few tries ๐งต๐ #CORL2026
Yiming Dou retweeted
Your navigation model already found the goal. It just doesn't know when to stop.
Introducing StopWatch, a navigation agent that grounds the goal explicitly before every action. It sets a new state of the art on six object-goal navigation benchmarks.
Watch the flower vase episode below. Both agents reach the goal. StopWatch stops there. LightNav-0 turns away and keeps going.
Code, models and data are open source ๐ stopwatch-nav.github.io/
Work done at Amazon FAR w/ @yunshengtian @QiuYuhengQiu @pabbeel @peterxichen @GuanyaShi @simonforrest
๐งต [1/n]
Dexterity from 0 sim & 0 RL! Check out our recent work DSP ๐๏ธ
We taught a robot to play!
Introducing Dexterity from Supervised Play (DSP): turning minutes of human play into a world model and dexterous manipulation policy. No task demonstrations or sim2real!
play-policy.github.io
Yiming Dou retweeted
Can we make dexterous robots compliant without sacrificing precise control?
Excited to introduce FOMO-Dex: Force-Modulated Dexterity! ๐๏ธ
FOMO-Dex brings force-aware compliance to dexterous teleoperation, enabling robots to pick up thin cards and chips ๐, cut paper ๐ช, turn nuts ๐ฉ, and use screwdrivers ๐ช.
These force-aware demonstrations also train better autonomous policies!
๐ฅ Videos & paper: yingyuan0414.github.io/fomo-โฆ
๐งต๐ (1/8)
Yiming Dou retweeted
What to remember in the past to actโspatial, temporal or motion details?
We often assume whatโs useful, but thatโs hard to know in advance.
Meet IMPRINT ๐ง : retain rich memory first, then learn to use it for control. Compact latents keep inference fast as history grows. ๐งต 1/10