Vladimir Yugay

Doing research in 3D Computer Vision.
Ph.D. student at @UvA_Amsterdam. Previously at @Meta @TU_Muenchen. Top 100 at @LegionTD2

[Vladimir Yugay @vyuga3d]

📽️ Check out Visual Odometry Transformer! VoT is an end-to-end model for getting accurate metric camera poses from monocular videos.
https://t.co/6tVXVt6mTx

VoT does not require camera calibration parameters, post-optimization, and operates in real-time, capable of processing thousands of frames. It is trained on a vast amount of real-world indoor data, but can work just fine in outdoor scenarios. It uses only camera poses as supervision - no optical flow, intrinsics, point clouds, or tracks - making it broadly accessible.

We experimented with different backbones, camera pose representations, scalability, and attention mechanisms. Our evaluation spans hundreds of full-length videos across various metrics, without aligning the predicted trajectory to the ground truth, to simulate a real-world application.

Thanks to the team, @kienduynguyen94, @theogevers, @cgmsnoek, and @Martin_R_Oswald from the @UvA_Amsterdam!

[Vladimir Yugay @vyuga3d]

@pablovelagomez1 @rerundotio @Gradio Nice work! Adding splatting-based slam like monogs would be great

[Vladimir Yugay @vyuga3d]

@techfusionjb It doesn't handle in view dynamics - only changes to the static scene itself. lots of work ahead

[Vladimir Yugay @vyuga3d]

⏩ GaME code release!
https://t.co/jnSgngCWJd
Grab components for your 3D reconstruction pipeline:
🔹 Purely geometric out-of-view scene change detection
🔹 Outdated observations filtering
🔹 Evaluation videos of changing scenes
Contributions welcome 🚀

[Vladimir Yugay @vyuga3d]

@gabriberton Didn't see much if any improvement for odometry

[Vladimir Yugay @vyuga3d]

@bercankilic We'd love to throw more data in! Alas academic compute is limited

[Vladimir Yugay @vyuga3d]

📽️ Check out Visual Odometry Transformer! VoT is an end-to-end model for getting accurate metric camera poses from monocular videos.
https://t.co/6tVXVt6mTx

VoT does not require camera calibration parameters, post-optimization, and operates in real-time, capable of processing thousands of frames. It is trained on a vast amount of real-world indoor data, but can work just fine in outdoor scenarios. It uses only camera poses as supervision - no optical flow, intrinsics, point clouds, or tracks - making it broadly accessible.

We experimented with different backbones, camera pose representations, scalability, and attention mechanisms. Our evaluation spans hundreds of full-length videos across various metrics, without aligning the predicted trajectory to the ground truth, to simulate a real-world application.

Thanks to the team, @kienduynguyen94, @theogevers, @cgmsnoek, and @Martin_R_Oswald from the @UvA_Amsterdam!

[Vladimir Yugay @vyuga3d]

@bercankilic Difficult. Lots of dynamics happening in the view - this is quite different from the data it was trained on. However, if tuned on egocentric dynamic data, I'm pretty confident it would work

[Vladimir Yugay @vyuga3d]

@liyuan_zz Ty!

[Vladimir Yugay @vyuga3d]

@bllchmbrs customer support is always online😂 ty!

[Vladimir Yugay @vyuga3d]

@bllchmbrs Barely have time to sleep these days 😓

[Vladimir Yugay @vyuga3d]

@gabriberton @changh95 @AjdDavison Exactly. Another aspect is that they were all relatively simple indoor scenes and we had input depth maps for extra verification

[Vladimir Yugay @vyuga3d]

@changh95 @AjdDavison For me even small dino features worked better than dbovw for loop closure detection, and there's quite some stuff from @gabriberton

[Vladimir Yugay @vyuga3d]

@gabriberton It's a very strong claim. It may be fine for robots, but in the wild reconstruction (e.g. random phone videos) slam (both dense and sparse) is very very far from being solved. It becomes even more obvious when moved a bit further from academic datasets

[Vladimir Yugay @vyuga3d]

@maikelborys No. We think that it will be easier for a person who knows ros to adapt it to ros, than for a person who doesn't know ros to adapt it to pure python :)

[Vladimir Yugay @vyuga3d]

⏩Code release for MAGiC-SLAM!
https://t.co/GHUbY2s54U

We vibe-coded hard to make the code as simple as possible. Here are some features you can seamlessly integrate into your 3D reconstruction pipeline right away:

[Vladimir Yugay @vyuga3d]

@kommentlezz Not really. Sfm is typically done offline and operates with 3D point clouds, not 3D gaussians.

[Vladimir Yugay @vyuga3d]

Introducing “Gaussian Mapping of Evolving Scenes”! We present an RGBD mapping system with novel view synthesis capabilities that accurately reconstruct scenes that change over time.
https://t.co/9xz7zvf0xx

[Vladimir Yugay @vyuga3d]

@AdamWHarley Absolutely love using the model

[Rerun @rerundotio]

One of our CVPR highlights 👉 Meet MAGiC-SLAM: multi-agent SLAM powered by rigidly deformable 3D Gaussians for novel view synthesis. New tracking, map-merge & loop-closure kill drift, align maps, and run faster + more accurately than 2-agent baselines on synthetic & real data. @vyuga3d @theogevers @Martin_R_Oswald #CVPR2025

[Vladimir Yugay @vyuga3d]

I will be presenting our previous work at CVPR Nashville. Drop by if you want to chat!

[Vladimir Yugay @vyuga3d]

This work was conducted in collaboration with Kersten Thies, @lucacarlone1 , @theogevers , @martinoswald , and Lukas Schmid at the Computer Vision Group of the @UvA_Amsterdam and @MIT Spark Lab.