💥 Geo4D: VideoGen 4D Scene 💥

The Oxford VGG unveils Geo4D, a breakthrough in #videodiffusion for monocular 4D reconstruction. Trained only on synthetic data, Geo4D still achieves strong generalization to real-world scenarios. It outputs point maps, depth, and ray maps, setting a new #SOTA in dynamic scene reconstruction. Code is now released!

⚡️ Review: https://t.ly/X55Uj
⚡️ Paper: https://arxiv.org/pdf/2504.07961
⚡️ Project: https://geo4d.github.io/
⚡️ Code: https://github.com/jzr99/Geo4D

#Geo4D #4DReconstruction #DynamicScenes #OxfordVGG #ComputerVision #MachineLearning #DiffusionModels

⚡️ BEST DATA SCIENCE CHANNELS ON TELEGRAM 🌟

🌼 SOTA Textured 3D-Guided VTON 🌼

👉 #ALIBABA unveils 3DV-TON, a novel diffusion model for HQ and temporally consistent video. Generating animatable textured 3D meshes as explicit frame-level guidance, alleviating the issue of models over-focusing on appearance fidelity at the expense of motion coherence. Code & benchmark to be released 💙

👉 Review: https://t.ly/0tjdC
👉 Paper: https://lnkd.in/dFseYSXz
👉 Project: https://lnkd.in/djtqzrzs
👉 Repo: TBA

#AI #3DReconstruction #DiffusionModels #VirtualTryOn #ComputerVision #DeepLearning #VideoSynthesis

https://t.iss.one/DataScienceT 🔗