Sequence-Based Machine Learning Model for Object Pattern Design

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional design systems for creating object patterns, such as those for turntable objects, are limited in their ability to provide users with control over design decisions and often result in unsatisfactory, repetitive, and mundane designs due to their reliance on pre-configured designs and retrieval approaches that fail to generate novel configurations.

Innovation Solution

A sequence-based machine-learning model is trained to analyze design characteristics from a set of training objects, generating state values for regions and recommending compatible design elements and characteristics for new objects, allowing users to select design elements and receive recommendations for stylistically compatible and aesthetically consistent designs.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of manufacture

If a retrieval approach using pre-configured designs is used, then the system is easy to implement, but the design capability is limited and produces repetitive results

Engineering Contradiction:

Improveease of implementationVSAvoiddesign capability

Core Design Contradiction:

Ease of manufactureVSAdaptability or versatility

Solution Approach 1:

The patent replaces the mechanical retrieval system (searching pre-configured designs in a database) with a sequence-based machine learning model that generates designs algorithmically. This substitution enables the system to move from static retrieval to dynamic generation, resolving the contradiction between ease of implementation and design capability.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent changes the fundamental parameter of design generation from fixed pre-configured options to dynamically generated sequences based on learned patterns. By training the model on design sequences and state values, the system can generate novel designs while maintaining stylistic consistency, thus improving adaptability without sacrificing implementability.

Inventive Principle:

Principle #35Parameter changes

2Ease of operation

If pre-configured designs are used, then the system is simple to operate, but users have limited control over design decisions

Engineering Contradiction:

Improvesimplicity of operationVSAvoiduser control

Core Design Contradiction:

Ease of operationVSAdaptability or versatility

Solution Approach 1:

The patent introduces dynamics into the design process by allowing users to provide input that influences the sequence-based model's generation. The system adapts to user input in real-time, generating designs that reflect user preferences while maintaining the simplicity of operation through automated processing.

Inventive Principle:

Principle #15Dynamics

3Ease of manufacture

If conventional design systems are used, then the implementation is straightforward, but novel design configurations cannot be produced

Engineering Contradiction:

Improveimplementation simplicityVSAvoiddesign repetitiveness

Core Design Contradiction:

Ease of manufactureVSObject-generated harmful factors

Solution Approach 1:

The patent replaces the mechanical retrieval system with a sequence-based machine learning model that generates designs algorithmically. This substitution enables the system to move from static retrieval to dynamic generation, resolving the contradiction between ease of implementation and design capability.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The model processes design generation in sequential steps, analyzing design characteristics, determining state values, and generating recommendations in a periodic sequence. This sequential processing enables novel configuration generation while maintaining implementation simplicity through structured computation.

Inventive Principle:

Principle #19Periodic action

4Device complexity

If retrieval approaches are used, then the system requires minimal complexity, but aesthetic consistency across regions cannot be ensured

Engineering Contradiction:

Improvesystem complexityVSAvoidaesthetic consistency

Core Design Contradiction:

Device complexityVSStability of the object's composition

Solution Approach 1:

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The model uses feedback from analyzed design characteristics and state values to generate aesthetically consistent recommendations. By learning from training data and adjusting generation based on input features, the system ensures aesthetic consistency without requiring complex manual configuration.

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS11093660B2Recommending pattern designs for objects using a sequence-based machine-learning model

Publication Date: 2021.08.17 ADOBE INC

AI summary

Methods and systems for aiding users in generating object pattern designs with increased speed. In particular, one or more embodiments train a sequence-based machine-learning model using training objects, each training object including a plurality of regions with a plurality of design elements. One or more embodiments identify a plurality regions of an object with a first region adjacent a second region. One or more embodiments receive a user selection of a design element for populating the first region with a first design element from a plurality of design elements. One or more embodiments identify a second design element from the plurality of design elements based on the first design element using the trained sequence-based machine-learning model. One or more embodiments also populate the second region with one or more instances of the second design element.