Загрузка видео...

Не удалось загрузить видео

Возникла проблема при загрузке этого видео. Это может быть связано с временными проблемами сети или видео может быть недоступно.

На главную

Today, we are announcing GroundedSLAM. LaMAria is the most difficult egocentric SLAM benchmark to date. We are the first to close the gap on Aria’s SLAM and eclipse all other methods by a wide margin. SLAM is the foundation of all 3D estimation. We think this is a big... show more

Grounded Superintelligence

1,284 subscribers

98,577 просмотров • 4 дней назад •via X (Twitter)

Наука и технологии

Anya Rossi• Live Now

Private livecam show

Комментарии: 7

Фото профиля Grounded Superintelligence

Grounded Superintelligence4 дней назад

GroundedSLAM is accurate locally and maintains its quality over long trajectories and difficult scenes. Other methods drift by >5m and deteriorate completely. Reliable SLAM is essential for robotics. Even slightly inaccurate head poses completely ruin mm-precise hand labels.

Фото профиля Grounded Superintelligence

Grounded Superintelligence4 дней назад

What GroundedSLAM gives you: - Single estimator = metric-scale from natural sensor fusion - Online calibration = sub-centimeter accuracy in manipulation settings - GPU-accelerated frontend for real-time estimation - Offline BA to correct estimator drift

Фото профиля Grounded Superintelligence

Grounded Superintelligence4 дней назад

Today, GroundedSLAM is unlocking applications across robotics, 3D scene reconstruction, real-to-sim, and world modeling by accurately estimating humans, objects, and environments.

Фото профиля Grounded Superintelligence

Grounded Superintelligence4 дней назад

Most importantly, our state-of-the-art egocentric SLAM is accessible to anyone today! Get a RoboCap and try it out with Grounded API at

Фото профиля BitRobot 🦾

BitRobot 🦾4 дней назад

Better SLAM = better robot training data. Congrats on releasing this for RoboCap users!

Фото профиля ThisMightWork

ThisMightWork4 дней назад

How much of the benchmark gain is already present in the online output, before offline correction? That split would help show where it fits: live robot control, training data, or both.

Фото профиля Nick Karpov

Nick Karpov4 дней назад

Do we need a cap for an invite code? How do we try

Похожие видео

We are publishing our first deep dive on what we believe is one of the most challenging layers in egocentric data - SLAM and VIO in the context of long-horizon state tracking. We break down how SLAM and VIO fail in egocentric settings - visual features vanish at close range, depth sensors saturate, fast head motion blurs frames, and these failures don't always occur in isolation. They hit at the exact same moment, leading to compounding errors and making the downstream data unusable. We believe the foundation for high-quality egocentric data demands sub-centimeter precision over long episodes ranging from a few minutes to up to an hour.

We are publishing our first deep dive on what we believe is one of the most challenging layers in egocentric data - SLAM and VIO in the context of long-horizon state tracking. We break down how SLAM and VIO fail in egocentric settings - visual features vanish at close range, depth sensors saturate, fast head motion blurs frames, and these failures don't always occur in isolation. They hit at the exact same moment, leading to compounding errors and making the downstream data unusable. We believe the foundation for high-quality egocentric data demands sub-centimeter precision over long episodes ranging from a few minutes to up to an hour.

FPV Labs

42,211 просмотров • 6 месяцев назад

Installing brembo brakes! 1st clips are “All Gas, Slam on the brakes” & how inefficient it is to stop, taking yards…Then we apply #gamespeed, learning to accelerate at a speed we can stop both for a Wide Base Crossover & Brake Plant Separate…Soon we will be taking turns w/ “no drift”

Installing brembo brakes! 1st clips are “All Gas, Slam on the brakes” & how inefficient it is to stop, taking yards…Then we apply #gamespeed, learning to accelerate at a speed we can stop both for a Wide Base Crossover & Brake Plant Separate…Soon we will be taking turns w/ “no drift”

Coach A Vogel

19,492 просмотров • 10 месяцев назад

We might sometimes disagree on policy, but we agree on the foundation of what America is: We are all equal and have the inalienable rights to life, liberty, and the pursuit of happiness, and the government of, by, and for the people is supposed to protect those rights.

We might sometimes disagree on policy, but we agree on the foundation of what America is: We are all equal and have the inalienable rights to life, liberty, and the pursuit of happiness, and the government of, by, and for the people is supposed to protect those rights.

Governor Josh Stein

10,552 просмотров • 7 месяцев назад

A GRAND SLAM FOR THE 2ND HALF CHAMPIONSHIP FOR JORGELYS MOTA!!!! WE ARE GOING BACK TO THE PLAYOFFS!!! OMGGGGGGGGGG

A GRAND SLAM FOR THE 2ND HALF CHAMPIONSHIP FOR JORGELYS MOTA!!!! WE ARE GOING BACK TO THE PLAYOFFS!!! OMGGGGGGGGGG

Fredericksburg Nationals

22,148 просмотров • 1 год назад

Who is Dr Dash? First of all this is Kigali today ♥️ You see why we Rwandans always stand by H.E. Paul Kagame? Can u smell the reason? Can you? I know you can😁 This is the legacy of President Paul Kagame. Peace, standards, development, accountability and most of all integrity Who am i? The person who believes that this is how we Rwandans are supposed to live. We deserve this, we really do. We still have a lot to work on, but we’ve laid a solid foundation. Proud Rwandan♥️

Who is Dr Dash? First of all this is Kigali today ♥️ You see why we Rwandans always stand by H.E. Paul Kagame? Can u smell the reason? Can you? I know you can😁 This is the legacy of President Paul Kagame. Peace, standards, development, accountability and most of all integrity Who am i? The person who believes that this is how we Rwandans are supposed to live. We deserve this, we really do. We still have a lot to work on, but we’ve laid a solid foundation. Proud Rwandan♥️

Dr. Dash

55,547 просмотров • 1 год назад

Look at the date. Still think we are wrong? Over a year ago we told you all this. John Jackson is one of the best accounts to follow on this platform. John Jackson

Look at the date. Still think we are wrong? Over a year ago we told you all this. John Jackson is one of the best accounts to follow on this platform. John Jackson

MAGA Cult Slayer

13,119 просмотров • 3 месяцев назад

TMC member and Kolkata Mayor Firhad Hakim last week: “Those who are not born in Islam are unfortunate. We have to bring them all of them under slam.“ The same man when the ‘fortunate ones’ are asked to reveal their names as sellers - Everyone is a human being first

TMC member and Kolkata Mayor Firhad Hakim last week: “Those who are not born in Islam are unfortunate. We have to bring them all of them under slam.“ The same man when the ‘fortunate ones’ are asked to reveal their names as sellers - Everyone is a human being first

Swati Goel Sharma

89,043 просмотров • 2 лет назад

Winston Peters puts a stake through Luxon’s heart this afternoon. The PM’s beloved Paris Climate agreement is a con. ‘Too many on the right are contaminated by seeping leftist group think, we are announcing a war on woke’ WP asked why we are sending money off overseas when U.S. Russia, China and India are creating all the pollution - NZ produces 0.17% of the worlds pollution. ‘We are a Nationalist party and this is a globalist agenda’ that’s why we want this re-evaluated. Luxon is going to have fun reconciling NZFirst’s stance.

Winston Peters puts a stake through Luxon’s heart this afternoon. The PM’s beloved Paris Climate agreement is a con. ‘Too many on the right are contaminated by seeping leftist group think, we are announcing a war on woke’ WP asked why we are sending money off overseas when U.S. Russia, China and India are creating all the pollution - NZ produces 0.17% of the worlds pollution. ‘We are a Nationalist party and this is a globalist agenda’ that’s why we want this re-evaluated. Luxon is going to have fun reconciling NZFirst’s stance.

Holyhekatuiteka

30,666 просмотров • 1 год назад

Jason Ma reveals the first true scaling law in robotics: robots trained on 1 million hours of first-person human video predictably get better without ever seeing robot data "Yesterday we announced our new flagship robot foundation model called Dyna-2. It's very significant for the entire field for many different reasons. First of all, it's the first robot foundation model trained on at least one million hours of data. That in itself is a very challenging infrastructure challenge." "What we demonstrated is that even if the one million hours of data is entirely first-person video of humans doing manipulation tasks, we actually see scaling law for transferring to robot embodiments that the model has never seen before." "By training on more human data, we actually see predictable performance improvement on robots. That's a big deal because the biggest challenge facing robot foundation models is that we don't have enough data." "Unlike language models, there are just not enough robotics data out there, and it's very difficult to collect these robot data. Only by passively observing humans doing things can we scale this kind of data, and this is the first time we showed there is a lot of improvement that can transfer." Jason Ma Dyna Robotics

Jason Ma reveals the first true scaling law in robotics: robots trained on 1 million hours of first-person human video predictably get better without ever seeing robot data "Yesterday we announced our new flagship robot foundation model called Dyna-2. It's very significant for the entire field for many different reasons. First of all, it's the first robot foundation model trained on at least one million hours of data. That in itself is a very challenging infrastructure challenge." "What we demonstrated is that even if the one million hours of data is entirely first-person video of humans doing manipulation tasks, we actually see scaling law for transferring to robot embodiments that the model has never seen before." "By training on more human data, we actually see predictable performance improvement on robots. That's a big deal because the biggest challenge facing robot foundation models is that we don't have enough data." "Unlike language models, there are just not enough robotics data out there, and it's very difficult to collect these robot data. Only by passively observing humans doing things can we scale this kind of data, and this is the first time we showed there is a lot of improvement that can transfer." Jason Ma Dyna Robotics

MTS

17,889 просмотров • 2 месяцев назад

The missing piece for the ChatGPT moment of robotics is finally here Today, we are releasing phosphobot, an open-source app to control robots intuitively and train the most powerful AI models for the cheapest robots possible -> phospho-app/phosphobot 🧵 What we are building:

The missing piece for the ChatGPT moment of robotics is finally here Today, we are releasing phosphobot, an open-source app to control robots intuitively and train the most powerful AI models for the cheapest robots possible -> phospho-app/phosphobot 🧵 What we are building:

Pierre-Louis Biojout (PLB)

35,860 просмотров • 1 год назад

Today we're announcing microSLAM, a monocular SLAM system built by our computer vision team in Zürich with ETH Zurich's Computer Vision and Geometry Lab. It ranks first on the LaMaria benchmark for monocular SLAM, and holds up against systems that carry many more cameras and IMUs. Ours runs on a single RGB stream. Robot learning is bottlenecked on data, and the largest untapped source is people going about their work. A camera on someone's head for an afternoon in a plant is a record of how that plant actually operates, including all the parts nobody writes down. That data is only useful if you can recover the geometry, and geometry from one moving camera in a world that will not sit still is the hard version of the problem. Solving it monocular is what makes the data cheap. Rigs do not scale to every worker. Glasses do. microSLAM turns that footage into interactive 3D environments where behavioral foundation models can be trained. One person walks a factory floor. A robot learns to navigate it.

Today we're announcing microSLAM, a monocular SLAM system built by our computer vision team in Zürich with ETH Zurich's Computer Vision and Geometry Lab. It ranks first on the LaMaria benchmark for monocular SLAM, and holds up against systems that carry many more cameras and IMUs. Ours runs on a single RGB stream. Robot learning is bottlenecked on data, and the largest untapped source is people going about their work. A camera on someone's head for an afternoon in a plant is a record of how that plant actually operates, including all the parts nobody writes down. That data is only useful if you can recover the geometry, and geometry from one moving camera in a world that will not sit still is the hard version of the problem. Solving it monocular is what makes the data cheap. Rigs do not scale to every worker. Glasses do. microSLAM turns that footage into interactive 3D environments where behavioral foundation models can be trained. One person walks a factory floor. A robot learns to navigate it.

Gianni Hodel

71,409 просмотров • 1 месяц назад

#WATCH | Varanasi, Uttar Pradesh: Megha Mukherjee, a BHU student, says, "For eight days, we have been sitting in front of the Central Office protesting for the removal of the Chief Proctor... There are 8-10 demands of ABVP. The main demand is the removal of the Chief Proctor, and various other demands... We don't want any person to go on the stage and slam our Chief Minister and the soldiers of the country, calling them rapist. We are not against the particular individual, but we are against the anti-national ideas she (Neha Bora, National President of the All India Students' Association) is carrying..."

#WATCH | Varanasi, Uttar Pradesh: Megha Mukherjee, a BHU student, says, "For eight days, we have been sitting in front of the Central Office protesting for the removal of the Chief Proctor... There are 8-10 demands of ABVP. The main demand is the removal of the Chief Proctor, and various other demands... We don't want any person to go on the stage and slam our Chief Minister and the soldiers of the country, calling them rapist. We are not against the particular individual, but we are against the anti-national ideas she (Neha Bora, National President of the All India Students' Association) is carrying..."

ANI

27,587 просмотров • 10 дней назад

Today, we are announcing an initiative to put NASA’s mission first and resource the objectives that matter most. We are organizing, planning, and investing in the talent and Centers to achieve the world-changing objectives only NASA is capable of accomplishing.

Today, we are announcing an initiative to put NASA’s mission first and resource the objectives that matter most. We are organizing, planning, and investing in the talent and Centers to achieve the world-changing objectives only NASA is capable of accomplishing.

NASA Administrator Jared Isaacman

550,799 просмотров • 4 месяцев назад

''These boys that we brought to the court today are all adults. Most of them are married men, None of them is a minor.'' — Counsel to the FG.

''These boys that we brought to the court today are all adults. Most of them are married men, None of them is a minor.'' — Counsel to the FG.

YabaLeftOnline

1,135,377 просмотров • 1 год назад

"In 2022, we made a bet. We said @Ethereum is all we need to build real-time, first-class DeFi products. Today we are proud to say that the MetaDEX protocols run on some of the most sophisticated DeFi apps that live entirely onchain."

"In 2022, we made a bet. We said @Ethereum is all we need to build real-time, first-class DeFi products. Today we are proud to say that the MetaDEX protocols run on some of the most sophisticated DeFi apps that live entirely onchain."

Dromos

35,278 просмотров • 10 месяцев назад