正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

Very impressive! At we already: learn from 3rd-person human videos + RL -- for locomotion. Excited to see where this path goes next!

Junyi Zhang

1,221 subscribers

17,377 次观看 • 1 年前 •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

2 条评论

FUTURE GAME STUDIO(Death Verse) 的头像

FUTURE GAME STUDIO(Death Verse)1 年前

Even if they do this, there is a difference in the scale of MRC, so in the end it will probably just be a demonstration. 知らんけど

PowerBeatsVR 的头像

PowerBeatsVR3 年前

VR fitness app PowerBeatsVR has just made its way to the official Meta Quest store! Grab it now and enjoy a fun workout anywhere, anytime, and without any subscription ever:

相关视频

We are moving from model controller to RL for locomotion. The policy is starting to look quite robust. Even more convinced now that legs are easier than wheels, given the software complexity has shrunk tremendously with RL.

We are moving from model controller to RL for locomotion. The policy is starting to look quite robust. Even more convinced now that legs are easier than wheels, given the software complexity has shrunk tremendously with RL.

Sankaet

133,130 次观看 • 10 个月前

Cade: "I'm still very excited about this series and where it's going to push this team to. There's a lot of things we can learn from this game that we'll move forward and be better."

Cade: "I'm still very excited about this series and where it's going to push this team to. There's a lot of things we can learn from this game that we'll move forward and be better."

Omari Sankofa II

50,592 次观看 • 3 个月前

We used RL to train models that create curated context from long documents for downstream use by agents. The models sometimes learn to invent their own abbreviations and shorthand. Optimizing with RL for downstream use produces very different artifacts from ordinary summaries: shorter, denser, creatively concise. We call these neural cheat-sheets.

We used RL to train models that create curated context from long documents for downstream use by agents. The models sometimes learn to invent their own abbreviations and shorthand. Optimizing with RL for downstream use produces very different artifacts from ordinary summaries: shorter, denser, creatively concise. We call these neural cheat-sheets.

Applied Compute

36,155 次观看 • 26 天前

A quick update from Sir Nick - very excited for everything to come over the next few weeks. See you all soon!

A quick update from Sir Nick - very excited for everything to come over the next few weeks. See you all soon!

Sir Nick Faldo

77,938 次观看 • 4 个月前

We’re onsite today at the U.S. Women's Open at Riviera. Very excited to watch this exacting test in person, and to see what kind of start Nelly gets off to.

We’re onsite today at the U.S. Women's Open at Riviera. Very excited to watch this exacting test in person, and to see what kind of start Nelly gets off to.

No Laying Up

35,627 次观看 • 1 个月前

After 2 years of building Artisan, I’m stepping down as CEO. Grateful for everything we’ve built and excited to see where the team goes from here.

After 2 years of building Artisan, I’m stepping down as CEO. Grateful for everything we’ve built and excited to see where the team goes from here.

Jaspar Carmichael-Jack

54,635 次观看 • 1 年前

We’re rooting for you and we support you every step of the way. Congratulations, Unc, so excited to see where this next journey leads you. 🎥: Walter Johnson | Mr Tendernism

We’re rooting for you and we support you every step of the way. Congratulations, Unc, so excited to see where this next journey leads you. 🎥: Walter Johnson | Mr Tendernism

Lee Merritt

13,407 次观看 • 4 个月前

my Luminosity Rising Stars Top 100 Introduction thank you to Luminosity Gaming & Cash App for the opportunity, excited to see where this goes 🙏 $ShiroRz

my Luminosity Rising Stars Top 100 Introduction thank you to Luminosity Gaming & Cash App for the opportunity, excited to see where this goes 🙏 $ShiroRz

shiro

14,897 次观看 • 1 年前

From DJ sets to deadlifts! We talked with LILJON on TMZ Live about his bodybuilding debut where he took home an impressive 3rd place title.

From DJ sets to deadlifts! We talked with LILJON on TMZ Live about his bodybuilding debut where he took home an impressive 3rd place title.

TMZ Live

31,222 次观看 • 10 个月前

gm from my hotel room where i’m figuring out how those 1-person videos work 😂 we are experimenting this week, messing around with types of videos we haven’t made before. we’ll see if we come up with any real+good versions with this 1-person format!

gm from my hotel room where i’m figuring out how those 1-person videos work 😂 we are experimenting this week, messing around with types of videos we haven’t made before. we’ll see if we come up with any real+good versions with this 1-person format!

Nevin Freeman

17,822 次观看 • 1 年前

At #ICRA2025, I'm about to present: "R+X: Retrieval and Execution from Everyday Human Videos". By using a VLM for retrieval and in-context IL for execution, robots can now learn from just videos of humans! Led by George Papagiannis and Norman Di Palo. See:

At #ICRA2025, I'm about to present: "R+X: Retrieval and Execution from Everyday Human Videos". By using a VLM for retrieval and in-context IL for execution, robots can now learn from just videos of humans! Led by George Papagiannis and Norman Di Palo. See:

Edward Johns

11,962 次观看 • 1 年前

Been pretty busy this spring but I’d like to thank God for my time at Tennessee and where he has my feet now. Tennessee was better than I could’ve imagined to pray for and I’m excited to see how this next chapter goes! Will always be a VFL but Boiler Up 🚂⬆️

Been pretty busy this spring but I’d like to thank God for my time at Tennessee and where he has my feet now. Tennessee was better than I could’ve imagined to pray for and I’m excited to see how this next chapter goes! Will always be a VFL but Boiler Up 🚂⬆️

Luke Poorman - LProductions

68,379 次观看 • 1 年前

Mobile-TeleVision follows our previous idea on upper-body and lower-body separation for control: 1⃣We train RL for lower-body locomotion and standing; 2⃣We use IK and retargeting directly for upper-body for high-precision and smooth manipulation. We avoid using RL for upper-body motion tracking. This is because fundamentally it is hard to avoid errors with RL tracking, and sim2real gap brings another gap. At this point, we still find using IK is more practical and precise.

Mobile-TeleVision follows our previous idea on upper-body and lower-body separation for control: 1⃣We train RL for lower-body locomotion and standing; 2⃣We use IK and retargeting directly for upper-body for high-precision and smooth manipulation. We avoid using RL for upper-body motion tracking. This is because fundamentally it is hard to avoid errors with RL tracking, and sim2real gap brings another gap. At this point, we still find using IK is more practical and precise.

Xiaolong Wang

32,662 次观看 • 1 年前

I'm very passionate about educating & helping others to protect themselves. As this video shows though I'm very bad at hyping my class where you can learn 1st hand from me on how to help others too.😁 See you at DEF CON

I'm very passionate about educating & helping others to protect themselves. As this video shows though I'm very bad at hyping my class where you can learn 1st hand from me on how to help others too.😁 See you at DEF CON

Jayson E. Street 💙 🤗💛

11,908 次观看 • 3 年前

Another impressive performance from 2-0 Imanol Rodriguez at Budo Sento. This kid goes for the kill #BudoSento15

Another impressive performance from 2-0 Imanol Rodriguez at Budo Sento. This kid goes for the kill #BudoSento15

caposa

31,128 次观看 • 3 年前