Custom 3D Object Texture Generation with Vertex-Based Neural Networks

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional texture rendering methods for 3D objects are limited, requiring pre-defined effects and not allowing users to create custom textures, leading to visible artefacts and limited options in AR/VR applications.

Innovation Solution

A method and system using vertex identification and texture generation-based neural networks to dynamically apply textures on 3D objects, generating texturing parameters including color and displacement vectors, and rendering a 2D image of the textured object.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If pre-defined texture effects are used in conventional solutions, then the rendering process is simple and fast, but users cannot create custom textures and effects, limiting versatility

Engineering Contradiction:
Improvetexture customization capabilityVSAvoidtexture generation system complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The system enables users to create their own custom textures and effects through neural network-based tools, eliminating the need for pre-defined effects. Users can generate unique textures by providing sketches or references, and the system automatically processes these into usable texture maps, allowing self-service texture creation without requiring manual manipulation of complex rendering parameters.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent replaces traditional manual texture creation methods (UV mapping, brush painting, parameter adjustment) with an AI-based neural network system. The neural networks automatically generate texture maps, normal maps, and other rendering parameters from user inputs, substituting the mechanical process of manual texture authoring with intelligent automated generation.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Manufacturing precision

If conventional UV-texture rendering pipeline is used, then the rendering process is straightforward, but artefacts are visible in multiple regions of the input 3D object

Engineering Contradiction:
Improvetexture mapping accuracyVSAvoidvisible artefacts on 3D object
Core Design Contradiction:
Manufacturing precisionVSObject-affected harmful factors

Solution Approach 1:

The system replaces the conventional UV-texture mapping pipeline with a neural network-based approach that directly generates high-quality texture maps without traditional UV unwrapping. This substitution eliminates artefacts by using learned patterns from training data to produce seamless, high-resolution textures that adapt to the 3D object's geometry without the distortions inherent in UV mapping.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent changes the fundamental parameters of texture generation by moving from fixed UV-coordinate-based mapping to flexible pixel-space generation guided by neural networks. This allows dynamic adjustment of texture resolution, pattern density, and geometric adaptation parameters, enabling artifact-free rendering by optimizing texture parameters specifically for each 3D object's characteristics.

Inventive Principle:
Principle #35Parameter changes

3Adaptability or versatility

If manual effort is required to create texture and normal maps, then custom textures can be created, but the process is time-consuming and labor-intensive

Engineering Contradiction:
Improvecustom texture creationVSAvoidtexture creation time
Core Design Contradiction:
Adaptability or versatilityVSLoss of time

Solution Approach 1:

The system enables automatic self-service texture creation where users simply provide a sketch or reference image, and the neural network automatically generates complete texture sets including color maps, normal maps, and other rendering parameters. This eliminates the time-consuming manual processes of UV unwrapping, brush painting, and parameter tuning, reducing texture creation from hours of manual work to minutes of automated generation.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent replaces manual texture authoring tools with an AI-based generation system that automatically creates textures from simple user inputs. The neural networks perform the complex tasks of texture synthesis, normal map generation, and coordinate system transformation that previously required skilled artists to manually execute, dramatically reducing the time and expertise required for custom texture creation.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

4Adaptability or versatility

If conventional texture rendering is used in AR/VR applications, then the implementation is simple, but the options for effects are limited

Engineering Contradiction:
Improveeffect options in AR/VRVSAvoidtexture generation system
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The system provides self-service texture generation capabilities specifically optimized for AR/VR applications, allowing users to create custom effects and textures directly within the application environment. Users can generate textures that adapt to real-world objects captured via camera or imported 3D models, enabling unlimited effect options without requiring pre-defined asset libraries or complex external authoring tools.

Inventive Principle:
Principle #25Self-service

Data Source

PatentUS20250272908A1System and method for applying a texture on a 3D object
Publication Date: 2025.08.28 SAMSUNG ELECTRONICS CO LTD
  • US20250272908A1 patent drawing
  • US20250272908A1 patent drawing
  • US20250272908A1 patent drawing

AI summary

A method for dynamically transferring a texture on a 3D object is disclosed. The method includes receiving a pattern associated with the texture to be applied onto the 3D object; identifying, using a vertex identification-based neural network, a set of vertices of a mesh of the 3D object; generating, using a texture generation-based neural network, respective texturing parameters for each vertex in the set of vertices based on the pattern, wherein the respective texturing parameters comprise a color vector and a displacement vector; generating a textured 3D object by applying the texture on the 3D object using the respective texturing parameters; and rendering a 2D image of the textured 3D object based on a rasterization of the textured 3D object.