Procedural Image Generation System for Latency and Content Control

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current image generation using deep learning neural networks faces issues such as lack of responsible AI, data and licensing legal concerns, and high latency, making it unsuitable for real-time applications and potentially generating inappropriate content.

Innovation Solution

A data processing system that includes a processor and a machine-readable medium storing executable instructions. The system receives textual inputs, analyzes them to predict color palettes, and uses procedural generation algorithms to create images while imposing limits on content and style, thereby ensuring appropriateness and reducing latency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If deep learning neural networks are used for image generation, then image generation capability is improved, but content appropriateness and legal compliance deteriorate

Engineering Contradiction:

Improveimage generation capabilityVSAvoidcontent appropriateness

Core Design Contradiction:

Adaptability or versatilityVSReliability

Solution Approach 1:

The patent extracts the harmful element (trained neural network model) from the image generation system and replaces it with a procedural generation algorithm. This extraction eliminates the risk of inappropriate content while preserving the core function of generating images from text descriptions, directly resolving the contradiction between image generation capability and content appropriateness.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent introduces an intermediary mechanism (procedural generation algorithm with content filters) between the text input and image output. This intermediary layer controls and monitors the generation process to ensure content appropriateness while maintaining the ability to generate diverse and creative images, thus resolving the reliability issue without sacrificing adaptability.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Adaptability or versatility

If deep learning neural networks are used for image generation, then image generation capability is improved, but latency increases

Engineering Contradiction:

Improveimage generation capabilityVSAvoidlatency

Core Design Contradiction:

Adaptability or versatilityVSLoss of time

Solution Approach 1:

The patent replaces the complex mechanical system (deep learning neural network) with a more efficient procedural generation algorithm. This substitution maintains the image generation capability while significantly reducing computational complexity and processing time, thereby eliminating the latency problem associated with neural network-based systems.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Manufacturing precision

If machine learning models are trained on protected data, then image generation quality is improved, but legal liability increases

Engineering Contradiction:

Improveimage generation qualityVSAvoidlegal liability

Core Design Contradiction:

Manufacturing precisionVSObject-affected harmful factors

Solution Approach 1:

The patent converts the potential harm (use of protected data in training) into a benefit by adopting a procedural generation approach that does not require training data. This method generates images through algorithmic processes rather than learning from existing datasets, thereby eliminating legal liability while maintaining or improving image generation quality through controlled procedural algorithms.

Inventive Principle:

Principle #22Blessing in disguise (Convert harm into benefit)

Data Source

PatentUS12266034B2Techniques for abstract image generation from multimodal inputs with content appropriateness considerations

Publication Date: 2025.04.01 MICROSOFT TECHNOLOGY LICENSING LLC

AI summary

A data processing system implements a receiving a textual input comprising a query for a first image. The data processing system also implements analyzing the textual input to determine a predicted color palette associated with a subject matter of the query; and procedurally generating the first image using the predicted color palette. Another implementation of the data processing system implements providing the textual input to a first machine learning model to obtain the first image, the first machine learning model being trained using a dataset comprising abstract imagery and analyzing the textual input using the first machine learning model to obtain the first image in response to receiving the textual input.