Generative Interior Design Using Graph Neural Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video game development tools lack an efficient method for generating and customizing interior designs of buildings, particularly in life-simulation games, where players want to create spaces according to their preferences.
Innovation Solution
A computer-implemented generative interior design method using graph neural networks to generate floor plans and layouts for video game buildings. This method involves a two-stage approach: first, generating a floor plan using a floor plan generator model, and second, creating a layout for each interior space using a layout generator model comprising one or more graph neural networks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual interior design tools are used in video games, then players can customize spaces, but the process is time-consuming and lacks efficiency
Solution Approach 1:
The patent replaces manual mechanical design operations with an automated system comprising a floor plan generator model and a layout generator model using graph neural networks. The system automatically generates floor plans from boundary data and creates detailed layouts with furniture placement, substituting the manual mechanical process of room-by-room design with intelligent automated generation while maintaining customization capabilities through user preferences and constraints.
2Productivity
If automated interior design generation is implemented, then design efficiency improves, but flexibility and customization options may be reduced
Solution Approach 1:
The patent implements dynamic customization where the automated generation system adapts to user preferences, budget constraints, and style selections. The floor plan generator and layout generator models process user inputs dynamically to produce customized results. Users can modify preferences, adjust budgets, and change style requirements, and the system regenerates designs accordingly, maintaining flexibility while preserving automation efficiency.
Solution Approach 2:
The system incorporates feedback mechanisms where user preferences, constraints, and selections are processed by the generation models to produce customized interior designs. The models learn from and adapt to user feedback, allowing iterative refinement of designs based on user requirements, thereby maintaining customization flexibility alongside automated efficiency.
3Manufacturing precision
If detailed layout generation for each interior space is performed, then design precision improves, but computational complexity increases
Solution Approach 1:
The patent segments the interior design generation process into two distinct stages: floor plan generation and layout generation. The floor plan generator model first creates the overall spatial division from boundary data, then the layout generator model using graph neural networks processes each interior space separately to generate detailed furniture layouts. This segmentation allows each model to specialize in specific tasks, achieving high precision while managing computational complexity through divided responsibility.
Solution Approach 2:
The patent introduces an intermediary representation where the floor plan generator creates a structured floor plan that serves as input for the layout generator. This intermediate floor plan structure organizes spatial information in a way that facilitates precise layout generation while reducing the computational burden on the layout generator model, as it receives pre-processed spatial divisions rather than raw boundary data.
Data Source
AI summary
This specification describes a computer-implemented generative interior design method. The method comprises obtaining input data comprising boundary data. The boundary data defines a boundary of an interior region of a video game building. A floor plan for the interior region of the video game building is generated. This comprises processing the input data using a floor plan generator model. The floor plan divides the interior region into a plurality of interior spaces. A layout for at least one of the plurality of interior spaces defined by the floor plan is generated by a layout generator model comprising one or more graph neural networks. The layout represents a configuration of one or more objects to be placed in the interior region.