VOICE DRIVEN 3D STATIC ASSET CREATION IN COMPUTER SIMULATIONS

Smart overview of the Invention

The patent application describes a method for generating 3D scenes from natural language descriptions, which can be either text or voice input. This process involves extracting relevant keywords related to asset attributes and their placement from the description, using these to generate a 2D image through a generative model, and then reconstructing 3D objects from the 2D image using another neural model. These 3D objects can be assembled to match the specified placement or generated directly using a 3D generative model.

Technical Field

The invention is rooted in computer technology, offering solutions that improve the creation of content within computer simulations, particularly in the realm of video games. It facilitates the creation of initial prototype assets for artists, allowing them to iterate on these assets efficiently. This technology is relevant for consumer electronics and computer game networks, enhancing the visual appeal and customization options in digital environments.

Methodology

The method involves receiving natural language input, which is processed by a neural network to render a 2D image of a computer simulation asset. This 2D image is then converted into a 3D asset that can be used in computer simulations. The input can specify locations or multiple objects, ensuring the generated 3D asset matches these specifications. Artists can modify these assets using a computer, and modifications can be input via a microphone.

Device and Apparatus

The described device includes a computer memory with instructions executable by a processor to convert 2D images into 3D assets and present them in simulations. An apparatus with a processor and output device is programmed to identify and convert 2D images into 3D assets for use in simulations. This technology can be integrated into various consumer electronics devices such as game consoles, VR and AR headsets, smart TVs, and mobile devices.

System Architecture

The system architecture includes server and client components connected over networks such as the Internet, intranets, or VPNs. These components exchange data securely, employing firewalls and other network infrastructure for reliability. The system can include devices like CE devices, game consoles, and portable computers, which operate on various operating systems. These devices are equipped with displays, speakers, input devices, and network interfaces to execute the described principles and facilitate the generation and modification of 3D assets.