Video yükleniyor...

Video Yüklenemedi

Ana Sayfaya Dön

These are not CGI. Reinforcement learning is so back. When operating on strings, it gives us o3. When operating on physical motors, it gives us a perfect humanoid backflip and a robot creature that out-maneuvers almost every animal on earth. RL is one of the only learning algorithms that...

356,697 görüntüleme • 1 yıl önce •via X (Twitter)

11 Yorum

Aleksa Gordić 🍿🤖 profil fotoğrafı
Aleksa Gordić 🍿🤖1 yıl önce

boston dynamics demo is controlled and we were capable of doing that for years now (e.g. their Atlas robot) but the thing that unitree robot is doing in the natural environment i find it hard to believe this is not cgi? this is a ChatGPT moment for robotics? why isn't this all over the timelines?

Digital Currency profil fotoğrafı
Digital Currency2 yıl önce

From 3D modeling to VR/AR development, our MSc in Metaverse program equips you with the technical skills to excel in the rapidly evolving digital world. Don't miss out—enroll today! #UNIC #MScMetaverse

AI Notkilleveryoneism Memes ⏸️ profil fotoğrafı
AI Notkilleveryoneism Memes ⏸️1 yıl önce

the age of the human generated reward function is over. the age of the AI generated reward function has come.

Abe Murray profil fotoğrafı
Abe Murray1 yıl önce

For folks close to this - how brittle is this? The issue with Boston Dynamics was always the blooper reel and the brittle nature of the reality behind the videos. Is reliability here 95%? 99%? 5x9's? Any insiders have a take on this?

Yuchen Jin profil fotoğrafı
Yuchen Jin1 yıl önce

Unitree is crazy

Chris Paxton profil fotoğrafı
Chris Paxton1 yıl önce

Do you have a source on that BD video being RL?

Shiraz Akmal profil fotoğrafı
Shiraz Akmal1 yıl önce

Reward function

Thanos Variant profil fotoğrafı
Thanos Variant1 yıl önce

That Unitree is so close to being an Autobot

Tianbao Xie profil fotoğrafı
Tianbao Xie1 yıl önce

"Give me a reward function, and I shall move the world." Time to scale on it!!!

H profil fotoğrafı
H1 yıl önce

How often do you have to replace those wheels?

Randall Briggs profil fotoğrafı
Randall Briggs1 yıl önce

What are the reward functions in these cases, would you say?

Benzer Videolar

This work makes a humanoid robot do simple parkour moves by looking with a depth camera and choosing the right move on the fly. The big deal is that it turns lots of small human moves into long, real-time robot behavior, without hand-coding every transition or retraining for each new course. A humanoid robot is usually good at steady walking, but it often fails when it has to do fast moves like jumping up, vaulting, or rolling, and then keep going to the next obstacle. The hard part is that you cannot easily collect training data for every possible obstacle shape, distance, and mistake, so robots end up learning a few moves that only work in a narrow setup. This work starts from short clips of real human parkour moves, like stepping over, vaulting, climbing, and rolling. It uses motion matching, which is basically a smart “pick the next clip that fits best right now” search, to stitch those short clips into a long, smooth plan that looks like a human doing a whole course. Then it trains a controller with reinforcement learning (RL), which means the robot learns by trial and error to copy that plan while staying balanced and not falling. After training separate expert controllers for different moves, it compresses them into 1 controller that uses only onboard depth sensing and a simple “go this fast in this direction” command. In real tests on a Unitree G1 humanoid, it can clear multiple obstacles in a row, adapt when obstacles get moved, and climb a wall up to 1.25m.

Rohan Paul

37,121 görüntüleme • 3 ay önce

When we say that the aggressor must not receive any reward for the war, so that peace can truly last, everyone must understand – these are not just words. Unfortunately, the world already had the chance to verify this twelve years ago. Russia’s war against Ukraine began with the occupation of our Crimea, and the world effectively turned a blind eye to this. The leaders back then showed little interest in the protests and resistance in Crimea and in Ukraine’s overall sentiment. The world advised Ukraine to stay silent. That is exactly why Putin came to believe he could wage a much larger war and confront the West even more aggressively. Now, every year on February 26 – the Day of Resistance to the Occupation of Crimea – we remember this global lesson, honouring those who did not stay silent and did not falter in the face of Russian aggression, and we insist that holding the aggressor accountable for war is one of the strongest security guarantees – one of the strongest prerequisites for lasting peace. I thank everyone around the world who supports us in this, participates in the work of our Crimea Platform and other international formats that remind the world about Crimea and the meaning of its occupation by Russia. I thank everyone who helps Ukraine resist Russian repression in Crimea, assists us in returning people from captivity, and prevents the occupying regime from consolidating its power. Russian presence on our peninsula serves only war, and nothing else. There must be peace, and therefore Crimea is Ukraine, and the world must always recognize this fact. Qırım evine qaytmalıdır! Glory to Ukraine!

Volodymyr Zelenskyy / Володимир Зеленський

153,934 görüntüleme • 3 ay önce