yearly reminder everything looks exponential from the middle of a sigmoid

RelitLRM: Generative Relightable Radiance for Large Reconstruction Models Project website is highly recommended and comes with a web viewer for every presented result. Links in the comments. Contributions (cited): • Novel transformer-based generative relighting architecture.

Cool results! Proud of being one of the core inventors of Adobe’s video generative tech!

Hate waiting 10 minutes for 3D GS to render your favorite indoor or outdoor scenes? ⏳ Our feed-forward solution, Long-LRM, cuts it down to just 1 second! ⚡️ With a straightforward mix of Mamba2 and transformer, it scales up to 32 high-res input images. arthurhero.github.io/projects/llrm/

No NeRF~ No 3DGS~ Our LVSM delivers state-of-the-art view synthesis quality from sparse-view inputs! Time to fully embrace scalable large transformers for building '3D-free' 3D models and apps! Amazing work by our intern Haian Jin and collaborators. 🔗haian-jin.github.io/projects/LVSM/

Use Camera Control to direct every shot with intention. Learn how with today's Runway Academy.

💡Open-Sourced Camera Control to direct every shot with intention with our recently proposed DimensionX! 🚩 It can also create any 3D/4D scenes with our controllable video gen. Project page: chenshuo20.github.io/DimensionX/ Code: github.com/wenqsun/Dimens… Arxiv: arxiv.org/abs/2411.04928

We've released our paper "Generating 3D-Consistent Videos from Unposed Internet Photos"! Video models like Luma generate pretty videos, but sometimes struggle with 3D consistency. We can do better by scaling them with 3D-aware objectives. 1/N page: genechou.com/kfcw

I'm building a new research lab Engineering Dept focusing on 4D computer vision and generative models. Interested in joining us as a PhD student? Apply to the Engineering program by Dec 3 🗓️ postgraduate.study.cam.ac.uk/courses/direct… ChatGPT's "portrait of my current life"👇 elliottwu.com

🚀 Introducing CAT4D! 🚀 CAT4D transforms any real or generated video into dynamic 3D scenes with a multi-view video diffusion model. The outputs are dynamic 3D models that we can freeze and look at from novel viewpoints, in real-time! Be sure to try our interactive viewer!

Cool work by Rundi!

Sharing something exciting we've been working on as a Thanksgiving gift: Diffusion Self-Distillation (DSD), which redefines zero-shot customized image generation using FLUX. DSD is like DreamBooth, but zero-shot/training-free. It works across any input subject and desired

Introducing MegaSaM! 🎥 Accurate, fast, & robust structure + camera estimation from casual monocular videos of dynamic scenes! MegaSaM outputs camera parameters and consistent video depth, scaling to long videos with unconstrained camera paths and complex scene dynamics!