International Edition
Latest News
Technology

Adobe Patents AI Tool for Generating Navigable 3D Worlds from Text

Adobe Patent Filings Reveal Text-to-3D Navigable Worlds and Expanded AI Tools Adobe is expanding its artificial intelligence capabilities beyond 2D image and video generation by filing patents for text-to-3D navigable world creation, creativebloq.com reported. The US Patent and…

Adobe Patents AI Tool for Generating Navigable 3D Worlds from Text

Adobe Patent Filings Reveal Text-to-3D Navigable Worlds and Expanded AI Tools

Adobe is expanding its artificial intelligence capabilities beyond 2D image and video generation by filing patents for text-to-3D navigable world creation, creativebloq.com reported. The US Patent and Trademark Office recently published a series of patent applications from the company detailing new concepts for 3D environment generation, single-photo camera angle manipulation, and automated material matching.

From Text Prompts to Navigable 3D Geometries

The most prominent patent filing outlines a method for turning text descriptions into explorable three-dimensional environments. When a user inputs a prompt such as “a foggy Japanese garden at dusk,” an unspecified AI model constructs an intermediate spherical 2D picture similar to a 360-degree view. A second AI process then ingests that 2D space to build the underlying 3D geometry.

The resulting environment allows users to move forward, backward, and view the space from multiple perspectives. Constructing 3D environments typically requires artists to manually build hundreds of individual assets and paint surface details using procedural tools. While Adobe already maintains a suite of 3D programs through its Substance collection—including Substance Painter and Substance 3D Modeler—its current tools lack integrated Firefly-powered generative AI features.

Adobe Patents AI Tool for Generating Navigable 3D Worlds from Text

Expanding Beyond Current Market Offerings

Adobe is entering a space currently occupied by early-stage tools like Blockade Labs’ Skybox AI, which generates 360-degree scenes from text prompts without true navigable environment capabilities for video games or virtual reality. Alongside the text-to-3D world generator, Adobe’s recent patent batch includes concepts for automatically finding matching materials across 3D models, rendering 3D garment previews, and generating new camera angles from a single photograph. Patent filings do not guarantee immediate product development, as companies frequently secure intellectual property to protect ideas from competitors and signal ongoing internal innovation.

About the author: Anika Shah - Technology

MSc in Computer Science, senior reporter. Anika focuses on AI ethics, cybersecurity, and emerging hardware—frequently moderating panels at CES and Web Summit. “Anika Shah decodes tech breakthroughs and startup disruption shaping tomorrow’s digital landscape.”