Or Litany

@orlitany

Assistant professor @TechnionLive and Sr. Research Scientist @NVIDIA | (novel) views are my own

orlitany.github.io Joined March 2018

857 Following 2.8K Followers 482 Posts

Latest posts 5

Today, kids, we'll learn how to track in 3D! 👨‍🏫

Sounds hard? It's actually quite simple:

3D tracks = 2D tracks + 3D lifting + ReLift

Phew, what a ReLift! 😮‍💨

Assaf Singer Verified account @assaf_singer Nov 2

🚀 Introducing ReLift: dense 3D tracking from monocular video.

Strong 2D trackers provide correspondences.
Geometry models produce depth & cameras.

We show how to combine them, repair 3D errors, and recover occluded motion for SOTA 3D tracking.

All it takes: one small corrector + off-the-shelf models.

🌐 Project: relift3d.github.io
@MarounKhriesh @NoamRot @yoni_kasten @orlitany

View quoted post on X

Or Litany @orlitany Oct 26

if you want to come post-docing with me at the Technion, here's your chance: azrielifoundation.org/fellows/intern…

azrielifoundation.org International Postdoctoral Fellowship - The Azrieli Foundation The Azrieli Fellows Program is an elite group of academics who cultivate a network of leading professionals in Israel and around the world.

Or Litany @orlitany Sep 2, 2025

Always wanted to come work with me and my students but lacked funding? You’re out of excuses 🙂
The Azrieli Foundation (@azrielifdn) is sponsoring visiting PhD positions (up to 4 months) and postdocs (up to 3 years).
azrielifoundation.org/fellows/intern…
View quoted post on X

Or Litany @orlitany Oct 22

Excited to share our NeurIPS 2026 paper (+code): Learnable Spectral Activations 📢

Prior INR work tunes which frequencies are available. LSA learns how to compose them. Gains on audio, images, NeRFs & more.
📄 arxiv.org/abs/2610.07419

Kudos to Tamir Shor who led it like a champ

Or Litany @orlitany Oct 21

Video models generate remarkably rich, scene-aware dynamics. But how do we extract that implicit knowledge into 3D animations we can actually use?

💡 Our answer: Jointly generate video and its corresponding 3D motion, ready to animate a rig!

Noam Rotstein Verified account @NoamRot Oct 21

🎬 Excited to share our paper: Joints, Camera, Action! (JCA)

JCA generates video & 3D motion together for any rig!

Video models generate motion you can watch. But what if you could use that motion to animate a 3D asset, with motion that responds to its surroundings?

research.nvidia.com/labs/par/proje…

🧵

View quoted post on X

Or Litany @orlitany Oct 10

VideoMDM got accepted to Neurips 🎸
See you in Sydney

Or Litany @orlitany Jun 12

Excited to share our new paper: VideoMDM 📢

We propose a principled framework for training 3D motion diffusion models (e.g. MDM), using only 2D supervision from monocular videos -- no 3D ground truth required.

Project: videomdm.github.io
Paper: arxiv.org/abs/2606.13364
🧵

View quoted post on X