Eryn Qian

Building spatial intelligence at @theworldlabs | ex-Meta Hyperscape @RealityLabs | CMU

[Eryn Qian @ErynQian]

🤯🥳🥹🍾🥂

[World Labs @theworldlabs]

We are excited to announce that World Labs is joining @AMD.

The research and technical breakthroughs we have achieved since our founding in 2024 have given us a clear vision for AI’s potential to solve problems in the spatial and physical world.

Accelerating the future of spatial and physical intelligence requires scaling our efforts, scaling our reach, and getting closer to the hardware.

[Keunhong Park @KeunhongP]

atlas world speedrun

[Eryn Qian @ErynQian]

@chrisoffner3d @Izard_of_Widd It can produce splats directly from the inference call

[Eryn Qian @ErynQian]

I'm behind the 3D side of this magic, turning them into Gaussian splats you can actually move through. I spend my days with my face pressed against Atlas's worlds. What looks glam from across the room has to stay sharp an inch from the wall. Still feeling surreal.

[World Labs @theworldlabs]

Introducing Atlas:

The world's first multimodal world model that generates image and video frames with pixel-perfect camera control and reconstructs them in 3D.

Model the world, move the camera, and simulate space & time.

[MTS @MTSlive]

World Labs co-founder @jcjohnss predicts a real-to-sim-to-real future where Atlas could train a robot for any new environment in minutes using just 5 photos from your phone:

"One is this notion of fully in-context learning. Maybe I've got a robotics foundation model, then I can demonstrate the robot once how a task should be performed, and that's enough for the robot to figure out."

"Another version is real to sim to real. Maybe I've got my pre-trained robotics foundation model, but I want to adapt it to this particular environment, like this studio, this table, moving this microphone around."

"There's a version where you could come in here and take just a couple casual videos with your phone or a couple images with your phone, use Atlas to reconstruct the space, and now stage all kinds of robotics interactions in this studio space."

"If you could lower the barrier to doing that, you could do it in five minutes to take five photos, upload it to Atlas, have it generate a simulation, maybe describe in natural language what kind of task you want the robot to do, have an agent go build that simulation for you, RL fine-tune your general purpose robotics foundation model, and now you've got a robot maybe in the span of a couple minutes that could come in and is perfectly adapted to this space."
@theworldlabs

[Eryn Qian @ErynQian]

@Izard_of_Widd Atlas is an omni model that already works with both 2D and 3D output modalities. In addition, we have a refinement pipeline to polish the 3D further to deliver the results you see in the blog post and will see as a future user

[Eryn Qian @ErynQian]

This is absolutely wild

[Ben Mildenhall @BenMildenhall]

I'm particularly excited by Atlas's ability to reconstruct scenes from a very small number of input image -- was able to create this flythrough of London's Natural History Museum by combining 3 input images I found from completely separate sources on Google Images

[Eryn Qian @ErynQian]

@aollivier82 Yep exactly! Our base in-house model has near-perfect consistency, which is why the resultant 3D reconstruction could be so clean and sharp. You're more than welcomed to request early access to Atlas from our blog post.

[Eryn Qian @ErynQian]

@aollivier82 We export the worlds in 3D gaussian splats. Unlike videos they're explicit structures. So yes there's perfect consistency no matter how many times you rotate 😃