Say less, do more. All in on immersive
Toronto
Joined December 2014
- Tweets4.4K
- Following1.6K
- Followers673
- Likes19.5K
Mike Dopsa retweeted
What if you could step inside a van Gogh painting?
We used Mundo to imagine new viewpoints from a single painting.
Take a look at this short demo.
Project: mundo-world-model.github.io/
Blog: huggingface.co/blog/Luo-Yiha…
Mike Dopsa retweeted
Render layers are a neat tool to have. We can create stark or opinionated lighting rigs with them. The bottom example has two spotlights with wide cones, each targeting a separate layer, allowing me to get the visual effect I was after without the light flowing behind the wall.
ALT A simple graphic comparing a regular point light with the usage of two spotlights and render layers
Mike Dopsa retweeted
Blender 5.3 supports gaussian-splat natively, just drop your .ply file in the scene!
I'm collecting in this thread some cool things you can do by combining it with geometry-nodes and basic shader
#b3d #gaussiansplat
Mike Dopsa retweeted
We released the code and model weights here: github.com/ayushjain1144/tra…. Have fun! 🙂
1/ We introduce TrackEverything: a 3D point tracker that tracks all points across all frames of long videos (1000+ frames).
Our key idea is to tie computation to unique 3D scene content rather than redundant 2D pixels in a video.
trackeverything.github.io
Mike Dopsa retweeted
Today I am open-sourcing my custom SwiftUI animation sequencing package that I developed to drive the animations in Letters. It can do all sorts of things!
github.com/jazzychad/Animati…
Mike Dopsa retweeted
📢OneCanvas (NeurIPS'26 Spotlight) Code Release📢
What if a 2D VLM could see the whole scene in one view?
We reproject features from every frame into a single occlusion-free canvas to make a stock VLM reason in 3D!
🔗 github.com/baranowskibrt/one…
🌐 baranowskibrt.github.io/onec…
📢 OneCanvas: 3D Scene Understanding via Panoramic Reprojection
We extract features from video frames and reproject them into one occlusion-free view of the whole scene that a 2D VLM reads just like a normal image. We can center this view on any viewpoint, including an agent's own pose for situated reasoning.
The same projection lets us create spatial training tasks with no human annotation, solvable only by reasoning over the 3D positions of real object features placed on an otherwise empty canvas.
The result is a stock 2D VLM that reasons in 3D, setting a new state of the art across spatial benchmarks at far less compute.
🌐 baranowskibrt.github.io/onec…
▶️ youtu.be/NIaHLB9gA7s
Great work by @baranowskibrt & @davech2y
Mike Dopsa retweeted
🦏 RHINO: Reconstructing Human Interactions with Novel Objects from Monocular Videos
From one moving-camera RGB video, we reconstruct a 3D human, an unseen manipulated 3D object, and the static 3D scene in one world frame
Video input, then 3D output. 👉 Code and data are public!
Mike Dopsa retweeted
Ok this is pretty insane. Astra side-loaded a custom app for my DJI RC2 controller that takes a live video stream from my Mini 5 Pro drone, streams it to a @roboflow Workflow running RF-DETR via our Serverless Video API, & overlays the predictions back on the controller’s screen.
Mike Dopsa retweeted
Preview: LichtFeld is slowly pivoting towards an end-to-end pipeline.
Where traditional SfM pipelines fail, our SLAM-based dense reconstruction runs in real time at 30 FPS and 1920 px.
Four minutes of capture condensed into one.
Dense point cloud and camera poses from a single monocular walk with a DJI Osmo Action 4.
Mike Dopsa retweeted
Apple is so freaking weird. They just casually drop 3D authoring into the preview app on macOS. I can sign annoying PDFs and render a scene with ray tracing for 3D graphics in the same app.
Mike Dopsa retweeted
THIS IS WHAT ALL VR HEADSETS SHOULD DO TO ALL YOUR FLAT GAMES
Holy shit IT WORKS!! One shotted with GPT6 Astra
I now have a fork of UEVR where I can play any UE game within a volumetric window!
This actually renders the game in 3D - meaning, two real game cameras, not some depth map bullshit. Everything looks right
This is an alternative (and more comfortable IMO) way to play most 2D->VR hacked games as there’s no motion sickness, and no need to redesign the UI and interaction systems
Needs a lot more polish before I publish it but OMG IT WORKS!
Mike Dopsa retweeted
For robotic simulation, Atlas reconstructs a space from just a few photos and generates the photorealistic RGB and depth data any robot's sensors would observe on any trajectory.
Robots can now be trained and tested in far more spaces. Until now, scanning spaces like these required expensive equipment and time-consuming capture.
