AI Image Generation System with Salience Map Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The current process of creating images for print products is inefficient, as salespersons often need to create multiple drafts to align with customer mental pictures, leading to increased man-hours and costs without clearly identifying customer preferences or correctable areas.
Innovation Solution
A system that utilizes an image generation AI to create images based on customer text input, allowing for efficient direction of design drafts that align with customer preferences, and includes a saliency analysis feature to help designers understand customer favorites and correctable areas.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a salesperson creates multiple image drafts manually to align with customer preferences, then the customer satisfaction improves, but the man-hours and costs increase
Solution Approach 1:
The patent replaces the manual mechanical process of image creation with an automated AI-based system. The image generation AI automatically creates images based on text descriptions, eliminating the need for manual image drafting by salespersons and reducing the time and cost while maintaining customer satisfaction.
Solution Approach 2:
The system enables self-service image generation where the AI automatically creates images from text descriptions without requiring manual intervention. The salience map analysis also automatically identifies important regions, allowing the system to serve itself in understanding and improving image quality according to customer preferences.
2Productivity
If an image generation AI is used to create images efficiently, then the design direction determination becomes faster, but the customer's specific preferences and correctable areas remain unclear
Solution Approach 1:
The patent introduces feedback through salience map analysis that automatically identifies important regions in the generated images. This feedback mechanism provides information about which areas are most significant to the customer, enabling the system to understand customer preferences and make targeted improvements without losing important information.
Solution Approach 2:
The patent uses visual representation through salience maps that highlight different regions of the image with varying intensities or colors. This allows the customer's preferences and important areas to be visually distinguished, providing clear information about which parts need attention or modification.
3Manufacturing precision
If multiple image iterations are performed between customer and designer, then the final image quality improves, but the time and cost burden increases
Solution Approach 1:
The patent performs preliminary action by using salience map analysis to pre-identify important regions and customer preferences before the actual image refinement process. This preliminary understanding allows the designer to focus on critical areas from the start, reducing the need for multiple iterative revisions and accelerating the path to high-quality results.
Data Source
AI summary
A non-transitory computer-readable storage medium stores a program causing a computer to perform: first reception of receiving first text data; first transmission of transmitting the first text data to an image generation AI apparatus; first image obtainment of obtaining, from the image generation AI apparatus, first image data for the first text data, the first image data being generated by the image generation AI apparatus; and external apparatus transmission of transmitting the first text data and the first image data associated with one another to an external apparatus.