RT @liusiqi42: This is finally out! 5 years ago Nicolas and Josh showed me a video of a single ant kicking a ball towards a goal and ... pr…

https://t.co/xo1Xx4Mwzh

RT @markus_with_k: Humanoid Soccer! We learned that RL combined with skill transfer, imitation learning and self-play can scale to competi…

Humanoid Soccer! We learned that RL combined with skill transfer, imitation learning and self-play can scale to competitive and collaborative behaviour in soccer ⚽️ with exceedingly high-dimensional embodiments. https://t.co/pk3iUscspK https://t.co/0SlO8

RT @DrJimFan: Meanwhile, sit back and enjoy World Cup! Science paper: https://t.co/vPNLersytP New Scientist Video: https://t.co/aojTGyWTuR

RT @DrJimFan: Meanwhile, sit back and enjoy World Cup! Science paper: https://t.co/vPNLersytP New Scientist Video: https://t.co/aojTGyWTuR

RT @DrJimFan: Meanwhile, sit back and enjoy World Cup! Science paper: https://t.co/vPNLersytP New Scientist Video: https://t.co/aojTGyWTuR

RT @DrJimFan: Meanwhile, sit back and enjoy World Cup! Science paper: https://t.co/vPNLersytP New Scientist Video: https://t.co/aojTGyWTuR

RT @DrJimFan: Meanwhile, sit back and enjoy World Cup! Science paper: https://t.co/vPNLersytP New Scientist Video: https://t.co/aojTGyWTuR

Meanwhile, sit back and enjoy World Cup! Science paper: https://t.co/vPNLersytP New Scientist Video: https://t.co/aojTGyWTuR

RT @ScienceMagazine: #ICYMI from @SciRobotics: Researchers have implemented a 3-step approach that helped AI humanoids learn low-level, mid…

RT @ScienceMagazine: #ICYMI from @SciRobotics: Researchers have implemented a 3-step approach that helped AI humanoids learn low-level, mid…

RT @ScienceMagazine: #ICYMI from @SciRobotics: Researchers have implemented a 3-step approach that helped AI humanoids learn low-level, mid…

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

RT @ScienceMagazine: #ICYMI from @SciRobotics: Researchers have implemented a 3-step approach that helped AI humanoids learn low-level, mid…

RT @ScienceMagazine: #ICYMI from @SciRobotics: Researchers have implemented a 3-step approach that helped AI humanoids learn low-level, mid…

Quite awesome! The title picture looks totally like sth from the Polish major league😁#ekstraklasa #DeepLearning

ScienceMagazine: #ICYMI from @SciRobotics: Researchers have implemented a 3-step approach that helped AI humanoids learn low-level, mid-level, and high-level skills to play virtual soccer. https://t.co/VURNZq6UB8 https://t.co/Cz7Rk3ZEh8

RT @ScienceMagazine: #ICYMI from @SciRobotics: Researchers have implemented a 3-step approach that helped AI humanoids learn low-level, mid…

RT @ScienceMagazine: #ICYMI from @SciRobotics: Researchers have implemented a 3-step approach that helped AI humanoids learn low-level, mid…

RT @ScienceMagazine: #ICYMI from @SciRobotics: Researchers have implemented a 3-step approach that helped AI humanoids learn low-level, mid…

#ICYMI from @SciRobotics: Researchers have implemented a 3-step approach that helped AI humanoids learn low-level, mid-level, and high-level skills to play virtual soccer. https://t.co/b4LhKCpeeI https://t.co/Go8Doj8tjX

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

RT @ScienceMagazine: #ICYMI from @SciRobotics: Researchers have implemented a 3-step approach that helped AI humanoids learn low-level, mid…

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

RT @ScienceMagazine: #ICYMI from @SciRobotics: Researchers have implemented a 3-step approach that helped AI humanoids learn low-level, mid…

RT @ScienceMagazine: #ICYMI from @SciRobotics: Researchers have implemented a 3-step approach that helped AI humanoids learn low-level, mid…

RT @ScienceMagazine: #ICYMI from @SciRobotics: Researchers have implemented a 3-step approach that helped AI humanoids learn low-level, mid…

ScienceMagazine: #ICYMI from @SciRobotics: Researchers have implemented a 3-step approach that helped AI humanoids learn low-level, mid-level, and high-level skills to play virtual soccer. https://t.co/VURNZq6UB8 https://t.co/DEpCiyKKoG

@MoonMarket_ The stock East. We moved 14% Today this stock has a $6 million market cap this stock is only $0.42 cents this is our chance to make some real money on this stock make it a meme stock buy now

RT @ScienceMagazine: #ICYMI from @SciRobotics: Researchers have implemented a 3-step approach that helped AI humanoids learn low-level, mid…

RT @ScienceMagazine: #ICYMI from @SciRobotics: Researchers have implemented a 3-step approach that helped AI humanoids learn low-level, mid…

#ICYMI from @SciRobotics: Researchers have implemented a 3-step approach that helped AI humanoids learn low-level, mid-level, and high-level skills to play virtual soccer. https://t.co/b4LhKCpM4g https://t.co/ACpddnXBrp

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

It would be so cool to let Kayra play soccer at RoboCup at some point in time.

RT @liusiqi42: This is finally out! 5 years ago Nicolas and Josh showed me a video of a single ant kicking a ball towards a goal and ... pr…

RT @TechXplore_com: Watch how an AI system learns to play #soccer from scratch @SciRobotics https://t.co/nK8lFIUzAa https://t.co/91XW7TcaVW

From motor control to team play in simulated humanoid football https://t.co/2Xsn5wxrTM

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

Super happy for everyone on the team this is finally out! Very hard challenge and lots of work went into this 👏 These little folks and their funny moves has been the highlight of many research progress updates 😊

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

This is essentially a move away from tabula rasa. This work is moving towards integration of priors to speed up solution exploration of a complex control problems in robotics such as controlling a robot in a real & noisy world. https://t.co/ZmxESbJRrI

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

面白い https://t.co/K7Uqn6WJGd https://t.co/rmDvX8PIj4

RT @daniel_kraft: A @DeepMind #AI learns to play #soccer using decades of match simulations ⚽️ It learned to skilfully control digital hu…

. @boltyboy

RT @daniel_kraft: A @DeepMind #AI learns to play #soccer using decades of match simulations ⚽️ It learned to skilfully control digital hu…

Researchers have implemented a 3-step approach that helped AI humanoids learn low-level, mid-level, and high-level skills to play virtual soccer. @DeepMind @ucl https://t.co/x683fCIDdJ https://t.co/3iejI8a2hV

A @DeepMind #AI learns to play #soccer using decades of match simulations ⚽️ It learned to skilfully control digital humanoid soccer players by working through decades' worth of soccer matches in just a few weeks. @SciRobotics https://t.co/4f37dl4tDH

RT @srlm: .@DeepMind apresenta "método geral para APRENDER controle motor integrado de alta dimensão e tomada de decisão coordenada de long…

RT @srlm: .@DeepMind apresenta "método geral para APRENDER controle motor integrado de alta dimensão e tomada de decisão coordenada de long…

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

RT @srlm: .@DeepMind apresenta "método geral para APRENDER controle motor integrado de alta dimensão e tomada de decisão coordenada de long…

RT @srlm: .@DeepMind apresenta "método geral para APRENDER controle motor integrado de alta dimensão e tomada de decisão coordenada de long…

.@DeepMind apresenta "método geral para APRENDER controle motor integrado de alta dimensão e tomada de decisão coordenada de longo prazo" e demonstra a coisa em agentes autônomos que aprenderam a JOGAR #FUTEBOL sozinhos: https://t.co/ozdnMrBLDa. mais hora

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

From motor control to team play in simulated humanoid football https://t.co/3lFZ98VbbQ

Mindblowing results in RL from DeepMind this week. 1) Learn a "general-purpose motor control" NPMP module by imitating motion capture 2) Learn much more complicated tasks like football with guidance from the NPMP Great summary + gifs on blog: https://t.c

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

RT @markus_with_k: Remember this guy? Turns out he can do a lot more than just walking! Please check out the humanoid soccer paper and vid…

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

RT @nikq: 機械学習で動きまで含めてサッカーができるようになるものなのだなあ、、、 2002年くらいにパルクールを機械学習モデルでモーション生成できたらいいよねという話をしてたのを思い出す。当時はせいぜいGAとかだったな

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

RT @TechXplore_com: Watch how an AI system learns to play #soccer from scratch @SciRobotics https://t.co/nK8lFIUzAa https://t.co/91XW7TcaVW

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

RT @nikq: 機械学習で動きまで含めてサッカーができるようになるものなのだなあ、、、 2002年くらいにパルクールを機械学習モデルでモーション生成できたらいいよねという話をしてたのを思い出す。当時はせいぜいGAとかだったな https://t.co/3edrKiZ1z9

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

fascinating that @amitkatwala's ⤵️ makes the actual research: https://t.co/yqpwu4Vyym more intelligible to layppl* and it's like there is indeed info of dif quality at every intersection of budget n time,,, *vast n inclusive category 🆘🫠🧮🍅

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

機械学習で動きまで含めてサッカーができるようになるものなのだなあ、、、 2002年くらいにパルクールを機械学習モデルでモーション生成できたらいいよねという話をしてたのを思い出す。当時はせいぜいGAとかだったな

DeepMind (https://t.co/yy9S3LY3c7) researchers have developed AI agents capable of playing virtual football. Read more here https://t.co/hSvTkMKzWJ #nft #nfts #CryptoNews #Cryptocurency #crypto https://t.co/YMxvM0TEaN

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

RT @liusiqi42: This is finally out! 5 years ago Nicolas and Josh showed me a video of a single ant kicking a ball towards a goal and ... pr…

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

ヒューマノイドフットボー From motor control to team play in simulated humanoid football https://t.co/SNTMUpiIrJ

Very cool work from our former lab member @sbohez!

Impressive!

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

RT @liusiqi42: This is finally out! 5 years ago Nicolas and Josh showed me a video of a single ant kicking a ball towards a goal and ... pr…

RT @liusiqi42: This is finally out! 5 years ago Nicolas and Josh showed me a video of a single ant kicking a ball towards a goal and ... pr…

sick

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

RT @liusiqi42: This is finally out! 5 years ago Nicolas and Josh showed me a video of a single ant kicking a ball towards a goal and ... pr…

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…

RT @DeepMind: Coordination is key to human intelligence, yet it's difficult to learn using hand-crafted rewards. New research in @SciRoboti…