4D Generative Models From in the Wild Videos
Summary
USPTO published patent application US20260099987A1 disclosing a method for generating 4D generative models from 2D video frames using 3D Gaussians and neural networks. The system processes sequential image frames to construct deformed 3D Gaussians with motion features, enabling creation of dynamic 4D scene representations from real-world video inputs. The application, filed October 7, 2025 under application number 19352319, names Koki Nagano, Kaiwen Jiang, Shalini De Mello, and Xueting Li as inventors.
What changed
USPTO published a patent application covering methods for constructing 4D generative models from 2D video sequences. The technology involves receiving and processing 2D image frames to generate 3D Gaussians, defining motion basis features, generating augmented images, constructing deformed 3D Gaussians with motion features, and rendering 2D images from the deformed Gaussians to train neural networks for 4D representation generation.\n\nSoftware developers, AI researchers, and companies developing AI-powered video processing, neural rendering, 3D/4D modeling, immersive content creation, augmented reality, or virtual reality applications should review the claims to understand potential IP implications. The patent's CPC classifications (G06T 15/08, G06N 3/08, H04N 21/816) indicate relevance across computer graphics, neural networks, and video processing technologies.
Archived snapshot
Apr 18, 2026GovPing captured this document from the original source. If the source has since changed or been removed, this is the text as it existed at that time.
4D GENERATIVE MODELS FROM IN THE WILD VIDEOS
Application US20260099987A1 Kind: A1 Apr 09, 2026
Inventors
Koki NAGANO, Kaiwen JIANG, Shalini DE MELLO, Xueting LI
Abstract
At least one embodiment for generating 4D generative models from in the wild videos includes receiving a first 2D image frame, processing the first 2D image frame to generate 3D Gaussians, defining a set of motion basis features from the 3D Gaussians, receiving a second 2D image frame, generating a plurality of augmented images based on the first 2D image frame and the second 2D image frame, processing the plurality of augmented images to generate a plurality of motion features, constructing deformed 3D Gaussians from the motion basis features and the motion features, generating a rendered 2D image from the deformed 3D Gaussians, and generating a 4D representation using a neural network trained based on the rendered 2D image and the second 2D image frame.
CPC Classifications
G06T 15/08 G06N 3/08 H04N 21/816
Filing Date
2025-10-07
Application No.
19352319
Related changes
Get daily alerts for USPTO Patent Applications - AI & Computing (G06N)
Daily digest delivered to your inbox.
Free. Unsubscribe anytime.
Source
About this page
Every important government, regulator, and court update from around the world. One place. Real-time. Free. Our mission
Source document text, dates, docket IDs, and authority are extracted directly from USPTO.
The summary, classification, recommended actions, deadlines, and penalty information are AI-generated from the original text and may contain errors. Always verify against the source document.
Classification
Who this affects
Taxonomy
Browse Categories
Get alerts for this source
We'll email you when USPTO Patent Applications - AI & Computing (G06N) publishes new changes.
Subscribed!
Optional. Filters your digest to exactly the updates that matter to you.