Image Quality Evaluation Using Prompt-Image Feature Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image quality evaluation methods based solely on image characteristics, such as color and sharpness, result in deviations from actual image quality, reducing accuracy.
Innovation Solution
A method and apparatus that utilize a neural network model to evaluate image quality by fusing image feature information, text feature information, and interactive feature information, allowing for multi-dimensional quality evaluation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If image quality evaluation is performed based solely on image characteristics (color, texture, sharpness), then the evaluation process is simple, but the accuracy of image quality evaluation deteriorates due to deviations from actual image quality
Solution Approach 1:
The patent merges image characteristics evaluation with prompt text quality evaluation into a unified quality assessment framework. The quality evaluation model integrates both image features (color, texture, sharpness) and text features (prompt quality, image-text consistency) to produce a comprehensive quality score, thereby resolving the contradiction between simplicity and accuracy.
Solution Approach 2:
The patent employs a composite evaluation approach that combines multiple evaluation dimensions (image characteristics, prompt text quality, image-text consistency) into a unified quality assessment. This composite methodology integrates diverse evaluation criteria to achieve more accurate and comprehensive image quality evaluation.
2Measurement precision
If multi-dimensional quality evaluation (image features, text features, interactive features) is implemented, then image quality evaluation accuracy is improved, but the device complexity increases
Solution Approach 1:
The quality evaluation model is designed with multi-functionality to handle diverse evaluation tasks. It can simultaneously evaluate image characteristics, prompt text quality, and image-text consistency within a single unified framework, reducing the need for separate evaluation systems and managing complexity through functional integration.
Solution Approach 2:
The system performs self-service by automatically integrating multiple evaluation dimensions without requiring manual intervention. The quality evaluation model autonomously processes image features, text features, and their interactions to generate comprehensive quality assessments, reducing operational complexity while maintaining high accuracy.
Data Source
AI summary
A method of image quality evaluation, an electronic device, and a storage medium are provided. The method includes: obtaining a target image to be evaluated, the target image beings generated based on a neural network model and a target prompt text; inputting the target image and the target prompt text to a target quality evaluation model, the target quality evaluation model performing quality evaluation, based on target image feature information corresponding to the target image, target text feature information corresponding to the target prompt text, and interactive feature information, the interactive feature information being obtained by fusing the target image feature information and the target text feature information; and determining a target quality evaluation result corresponding to the target image based on an output of the target quality evaluation model.