ML Parameter Adjustment for Visual Content Context

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional content editing systems require manual and time-consuming adjustments of visual parameters, often resulting in inconsistent styles and inefficient use of device resources, as they fail to account for contextual information and do not provide mechanisms for users to adjust poor visual results.

Innovation Solution

An automatic digital parameter adjustment system that uses machine learning to predict parameter values based on insights learned from an image set, allowing for balanced visual output across various scenes and settings, without requiring access to the creation parameters or tools used in the corrected image.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual parameter adjustment is used, then users can control visual characteristics, but the process becomes time-consuming and results in inconsistent styles

Engineering Contradiction:

Improvevisual parameter control precisionVSAvoidimage editing time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system automatically adjusts digital image parameters by analyzing visual and contextual features without requiring manual user input. The machine learning model self-services the parameter adjustment task by predicting optimal values based on learned patterns from training data, eliminating the time-consuming manual adjustment process while maintaining consistent style across images

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system performs preliminary learning during a training phase where the machine learning model is trained on labeled image data with ground truth parameter values. This preliminary action enables the system to automatically predict appropriate parameter adjustments for new images without manual intervention, resolving the contradiction between precision control and time efficiency

Inventive Principle:

Principle #10Preliminary action

2Adaptability or versatility

If conventional adjustment systems are used, then basic parameter changes are possible, but contextual information is not considered leading to poor visual results

Engineering Contradiction:

Improveparameter adjustment flexibilityVSAvoidvisual result quality

Core Design Contradiction:

Adaptability or versatilityVSReliability

Solution Approach 1:

The system analyzes local visual features and contextual information within specific regions of the image to determine appropriate parameter adjustments. By considering local characteristics such as lighting conditions, scene type, and object properties, the system adapts parameter values to match the specific context of each image region, improving visual result quality while maintaining flexibility

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The machine learning model acts as an intermediary between the input image and the parameter adjustment process. It processes visual and contextual features through learned patterns and predictions to determine optimal parameter values, mediating the complex relationship between image characteristics and appropriate adjustments, thereby improving reliability without sacrificing adaptability

Inventive Principle:

Principle #24Intermediary (Mediator)

3Ease of operation

If manual parameter adjustment is performed, then users can achieve desired visual effects, but device resources are consumed through repetitive adjustments

Engineering Contradiction:

Improvevisual effect achievementVSAvoiddevice resource consumption

Core Design Contradiction:

Ease of operationVSUse of energy by moving object

Solution Approach 1:

The system performs automatic parameter adjustment without requiring repeated manual user actions. The machine learning model processes each image independently and predicts optimal parameters in a single operation, eliminating the energy-consuming cycle of manual adjustment attempts and retries, while still achieving the desired visual effects

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system replaces the mechanical interaction of manual slider adjustments and iterative parameter changes with an automated machine learning-based prediction system. This substitution eliminates the repetitive mechanical operations that consume device resources, while maintaining the ability to achieve visually pleasing effects through intelligent prediction

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS11930303B2Automated digital parameter adjustment for digital images

Publication Date: 2024.03.12 ADOBE INC

AI summary

Systems and techniques for automatic digital parameter adjustment are described that leverage insights learned from an image set to automatically predict parameter values for an input item of digital visual content. To do so, the automatic digital parameter adjustment techniques described herein captures visual and contextual features of digital visual content to determine balanced visual output in a range of visual scenes and settings. The visual and contextual features of digital visual content are used to train a parameter adjustment model through machine learning techniques that captures feature patterns and interactions. The parameter adjustment model exploits these feature interactions to determine visually pleasing parameter values for an input item of digital visual content. The predicted parameter values are output, allowing further adjustment to the parameter values.