Display Apparatus Automatic Style Transfer Using CNN Models

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing display apparatuses require significant user intervention to select the most optimized style for an input image, making the style transfer process time-consuming and cumbersome.

Innovation Solution

The display apparatus employs convolutional neural network (CNN)-based models corresponding to various styles, allowing for automatic selection and application of the most optimized style to an input image, minimizing user intervention.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If a user manually selects styles one by one to find the optimized style for an input image, then the user can identify the best matching style, but the process becomes time-consuming and cumbersome

Engineering Contradiction:

Improvestyle matching accuracyVSAvoidtime for style selection

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system performs automatic style selection without requiring user intervention. The server independently analyzes the input image, determines appropriate styles based on image attributes and environment information, and applies the optimized style automatically, eliminating the need for manual trial-and-error selection by the user

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system pre-processes the input image by analyzing its attributes and comparing them with stored style information before the user sees the result. This preliminary analysis allows the server to pre-select the most suitable style, so when the styled image is presented to the user, it is already optimized rather than requiring post-processing adjustment

Inventive Principle:

Principle #10Preliminary action

2Extent of automation

If a fixed method is used for style transfer, then the process is simple and automated, but the style may be unoptimized or mismatched for the specific input image

Engineering Contradiction:

Improveautomation of style transferVSAvoidstyle optimization accuracy

Core Design Contradiction:

Extent of automationVSMeasurement precision

Solution Approach 1:

The system applies different processing approaches based on local characteristics of the input image. By analyzing specific attributes of the input image (such as content type, color scheme, composition) and matching them with corresponding style information from the database, the system selects the most appropriate style for that specific image rather than applying a universal fixed method

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The system dynamically adjusts style selection based on varying parameters of the input image and environment. By changing the selection criteria according to image attributes (resolution, format, content) and environmental context (display device, viewing conditions), the system optimizes the style match for each specific case rather than using a static fixed method

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentEP3770860B1Style transfer of an image based on information related to a use environment

Publication Date: 2025.03.05 SAMSUNG ELECTRONICS CO LTD

Figure 1
Figure 2
Figure 3

AI summary

A display apparatus includes a display and a processor. The processor is configured to obtain information related to a use environment of the display apparatus, obtain a first image, identify a style among a plurality of styles which are applicable to the first image, based on the information related to the use environment, obtain a second image which is converted from the first image based on information related to the identified style, and control the display to display the second image.