Loading video...

Video Failed to Load

Go Home

Collecting DAgger style data for an ACT policy trained on 50 episodes - just for a minimal experiment. Problem: The parts hard for the policy are also hard for the human with teleoperation.

DominiqueCAPaul's profile picture

Dominique Paul

6,928 subscribers

11,223 views • 2 months ago •via X (Twitter)

LIVE

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

Another reason I’m interested in UMI style grippers is that tasks that‘d take 10s for a human take 5x as long with teleoperation. The lack of haptic and force feedback makes it hard to do movements relying on them like here, passing an object from one hand to another. Adding to that: teleoperation means you’re standing a meter away from the objects you’re operating on, and here for example, the box is obstructing the view of the PCBs. Thank you for great teleoperation Felix Neumann 🤝
0:46

Another reason I’m interested in UMI style grippers is that tasks that‘d take 10s for a human take 5x as long with teleoperation. The lack of haptic and force feedback makes it hard to do movements relying on them like here, passing an object from one hand to another. Adding to that: teleoperation means you’re standing a meter away from the objects you’re operating on, and here for example, the box is obstructing the view of the PCBs. Thank you for great teleoperation Felix Neumann 🤝

Dominique Paul

15,434 views • 7 months ago

🚨BREAKING: Marco Rubio just CLAPPED BACK HARD at Dem Sen. Jacky Rosen: ROSEN: “Are you even the lead for US government on foreign policy?!” RUBIO: “NO, the lead on US foreign policy is named Donald J. Trump, the President of the United States!”
0:56

🚨BREAKING: Marco Rubio just CLAPPED BACK HARD at Dem Sen. Jacky Rosen: ROSEN: “Are you even the lead for US government on foreign policy?!” RUBIO: “NO, the lead on US foreign policy is named Donald J. Trump, the President of the United States!”

Gunther Eagleman™

595,734 views • 6 months ago

The problem isn’t just with non-ethanol compliant vehicles. Even ethanol-compliant vehicles are now facing “uninvited guests” after this forced #Ethanol push. Before calling it a success story, Nitin Gadkari MUST answer why common citizens are paying the price for a policy experiment?
0:43

The problem isn’t just with non-ethanol compliant vehicles. Even ethanol-compliant vehicles are now facing “uninvited guests” after this forced #Ethanol push. Before calling it a success story, Nitin Gadkari MUST answer why common citizens are paying the price for a policy experiment?

ज़िद्दी नागरिक

79,403 views • 1 month ago

Age verification laws for porn sites are not just an attempt to make porn illegal, it's also an attempt to normalize the government collecting your data.
1:38

Sensitive content

Age verification laws for porn sites are not just an attempt to make porn illegal, it's also an attempt to normalize the government collecting your data.

Red Umbrella Club

69,208 views • 1 month ago

The problem with humanoid teleoperation is that it is expensive and difficult to scale Enter NVIDIA's EgoScale: - A VLA model pretrained on thousands hours of egocentric human videos. - Mid-trained via 50 hours of human + 4 hours of robot "play" data for human-robot alignment. - Fine-tuned with very few examples of task-specific robot teleoperation (100 or fewer per task). - Successfully transfers across 5-finger (Sharpa) and 3-finger (Unitree G1) robot hands. - Performance scales predictably as data increases.
0:50

The problem with humanoid teleoperation is that it is expensive and difficult to scale Enter NVIDIA's EgoScale: - A VLA model pretrained on thousands hours of egocentric human videos. - Mid-trained via 50 hours of human + 4 hours of robot "play" data for human-robot alignment. - Fine-tuned with very few examples of task-specific robot teleoperation (100 or fewer per task). - Successfully transfers across 5-finger (Sharpa) and 3-finger (Unitree G1) robot hands. - Performance scales predictably as data increases.

The Humanoid Hub

44,441 views • 5 months ago

For many Democrats, Israel “is not even so much a policy issue anymore,” says Erick Erickson. “It’s an emotional gut check for a lot of Democratic primary voters. And that becomes the problem when it goes beyond reason and public policy to emotion. It’s hard for the Democrats to then shift back towards Israel.” #Israel Erick Erickson Mark Halperin Kevin Walling 2WAY
5:29

For many Democrats, Israel “is not even so much a policy issue anymore,” says Erick Erickson. “It’s an emotional gut check for a lot of Democratic primary voters. And that becomes the problem when it goes beyond reason and public policy to emotion. It’s hard for the Democrats to then shift back towards Israel.” #Israel Erick Erickson Mark Halperin Kevin Walling 2WAY

2WAY

21,327 views • 1 month ago

We are aware of the immense value of data in an AI powered world. Assam State Data Policy 2026 strikes a fine balance between privacy and using anonymized data for better policy outcomes, informed choices and drives citizen centric governance.
0:48

We are aware of the immense value of data in an AI powered world. Assam State Data Policy 2026 strikes a fine balance between privacy and using anonymized data for better policy outcomes, informed choices and drives citizen centric governance.

Himanta Biswa Sarma

30,191 views • 1 month ago

Age verification laws for p*rn sites are not just an attempt to make p*rn illegal, it’s also an attempt to normalize the government collecting your data.
1:34

Sensitive content

Age verification laws for p*rn sites are not just an attempt to make p*rn illegal, it’s also an attempt to normalize the government collecting your data.

Adam Conover

140,251 views • 9 months ago

Presenting DemoDiffusion: An extremely simple approach enabling a pre-trained 'generalist' diffusion policy to follow a human-demonstration for a novel task during inference One-shot human imitation *without* requiring any paired human-robot data or online RL 🙂 1/n
0:30

Presenting DemoDiffusion: An extremely simple approach enabling a pre-trained 'generalist' diffusion policy to follow a human-demonstration for a novel task during inference One-shot human imitation *without* requiring any paired human-robot data or online RL 🙂 1/n

Homanga Bharadhwaj

32,919 views • 1 year ago

A shift is underway on Capitol Hill, with stablecoins leading the way Lily Liu on how the GENIUS Act signals a new era for U.S. crypto policy
1:04

A shift is underway on Capitol Hill, with stablecoins leading the way Lily Liu on how the GENIUS Act signals a new era for U.S. crypto policy

Solana

92,439 views • 1 year ago

.House Republicans just PASSED the One Big Beautiful Bill Act and delivered historic America First policy wins for the American people. On to the Senate!
1:55

.House Republicans just PASSED the One Big Beautiful Bill Act and delivered historic America First policy wins for the American people. On to the Senate!

Tom Emmer

33,979 views • 1 year ago

Robotics is still data starved. Collecting high-quality robot demonstrations remains brutally slow and expensive. Introducing COBALT: A cloud-native teleoperation platform designed for large-scale robot learning. We are democratizing data collection by leveraging the hardware everyone already owns: the smartphone All you need is to download an app (today)! Read on for more!
0:39

Robotics is still data starved. Collecting high-quality robot demonstrations remains brutally slow and expensive. Introducing COBALT: A cloud-native teleoperation platform designed for large-scale robot learning. We are democratizing data collection by leveraging the hardware everyone already owns: the smartphone All you need is to download an app (today)! Read on for more!

Animesh Garg

103,993 views • 2 months ago

SECRETARY RUBIO: "The President's policy is a head for an eye."
1:16

SECRETARY RUBIO: "The President's policy is a head for an eye."

Department of State

543,166 views • 21 days ago

Sec. Rubio on Iran: I saw their foreign minister say something about their policy is an eye for an eye. The President's policy is a head for an eye—honestly. That's what it's going to be. They're paying a very heavy price for the things they're doing.
1:10

Sec. Rubio on Iran: I saw their foreign minister say something about their policy is an eye for an eye. The President's policy is a head for an eye—honestly. That's what it's going to be. They're paying a very heavy price for the things they're doing.

Breaking911

68,551 views • 21 days ago

So you’ve trained your favorite diffusion/flow based policy, but it’s just not good enough 0-shot. Worry not, in our new work DSRL - we show how to *steer* pre-trained diffusion policies with off-policy RL, improving behavior efficiently enough for direct training in the real world! DSRL retains nice exploration from the base policy, but allows for quick improvement beyond this base policy with RL. The method is frustratingly simple, and super easy to throw on top of your favorite pretrained policy (VLA/diffusion policy, etc). Let’s think about how it works, 🧵 (1/10)
2:53

So you’ve trained your favorite diffusion/flow based policy, but it’s just not good enough 0-shot. Worry not, in our new work DSRL - we show how to *steer* pre-trained diffusion policies with off-policy RL, improving behavior efficiently enough for direct training in the real world! DSRL retains nice exploration from the base policy, but allows for quick improvement beyond this base policy with RL. The method is frustratingly simple, and super easy to throw on top of your favorite pretrained policy (VLA/diffusion policy, etc). Let’s think about how it works, 🧵 (1/10)

Abhishek Gupta

19,098 views • 1 year ago

TWIST: a real-time teleoperation system for humanoid robots to mimic whole-body motions. Reference motion data is generated by retargeting human motion-capture data to the robot. Then, the controller is trained in simulation using reinforcement learning and behavior cloning.
0:58

TWIST: a real-time teleoperation system for humanoid robots to mimic whole-body motions. Reference motion data is generated by retargeting human motion-capture data to the robot. Then, the controller is trained in simulation using reinforcement learning and behavior cloning.

The Humanoid Hub

46,715 views • 1 year ago

“It's just been hard for a really long time...” Abby Jitendra, Policy Adviser at the Joseph Rowntree Foundation, underscores why a new poll shows that for many people the cost of living remains their top concern. #Newsnight
1:14

“It's just been hard for a really long time...” Abby Jitendra, Policy Adviser at the Joseph Rowntree Foundation, underscores why a new poll shows that for many people the cost of living remains their top concern. #Newsnight

BBC Newsnight

21,212 views • 11 months ago