Simultaneous Image Expansion Without Text Prompts

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional image processing tools, such as Dall-E and Stable Diffusion, require manual text prompts and involve tedious tiling operations to expand images, which can be cumbersome for users.

Innovation Solution

An image processing apparatus and method that allows users to expand multiple sides of an input image simultaneously using an image generation model, without the need for text prompts, by generating a modified image with original content in one region and generated content in another region.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If conventional image processing tools (Dall-E, Stable Diffusion) are used to expand images, then image generation capability is achieved, but the process requires manual text prompts and tedious tiling operations

Engineering Contradiction:

Improveimage expansion operationVSAvoidtime required for image expansion

Core Design Contradiction:

Ease of operationVSLoss of time

Solution Approach 1:

The image expansion process is segmented into distinct functional regions: a first region containing original image content and a second region for generated content. This segmentation allows the system to process different areas with different requirements simultaneously, eliminating the need for iterative tiling operations while maintaining operational simplicity and reducing time loss.

Inventive Principle:

Principle #1Segmentation

2Productivity

If multiple tiling operations are performed to expand images on multiple sides, then complete image expansion is achieved, but the number of steps and complexity increases

Engineering Contradiction:

Improveimage expansion efficiencyVSAvoidnumber of expansion steps

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

Multiple image expansion operations that would traditionally be performed separately on different sides of an image are merged into a single simultaneous operation. The system processes expansion on multiple sides at once by defining a frame with multiple regions, thereby improving productivity while reducing the complexity of the expansion process from multiple sequential steps to one unified operation.

Inventive Principle:

Principle #5Merging (Combining)

3Ease of operation

If text prompts are required for image generation, then controlled content generation is achieved, but the ease of use and speed of operation decreases

Engineering Contradiction:

Improveimage generation operationVSAvoidtime for text prompt input

Core Design Contradiction:

Ease of operationVSLoss of time

Solution Approach 1:

The image generation system performs self-service by automatically generating content for the second region based on the original image content and frame definition, without requiring manual text prompts from the user. This eliminates the time loss associated with prompt input while maintaining controlled and coherent content generation through the model's inherent understanding of the original image.

Inventive Principle:

Principle #25Self-service

Data Source

PatentUS20250124626A1Generative expand in image editing applications

Publication Date: 2025.04.17 ADOBE INC

AI summary

A method, apparatus, non-transitory computer readable medium, and system for image generation include obtaining, via a user interface, an input image and a user input that indicates a frame for modifying the input image including a first region inside of the input image and a second region outside of the input image, and excluding a third region inside of the input image. A modified image is generated using an image generation model. The modified image includes original content from the input image in the first region and generated content in the second region, and excluding content from the input image in the third region. The modified image is presented for display in the user interface.