Intelligent Engine Text-to-Image Search Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current text-to-image search systems rely solely on similarity comparisons, leading to inaccurate search results due to the lack of a stable threshold for determining image matches.
Innovation Solution
The proposed method utilizes an intelligent engine with two distinct models: a first model for text-to-image matching and a second model for image-to-image matching, allowing for a refined search process by iteratively filtering images based on character information and image features.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If text-to-image search relies solely on similarity comparisons, then the search process is simple, but the search result accuracy is poor
Solution Approach 1:
The search process is segmented into multiple stages: first model for initial image set selection, second model for refined matching. This divides the complex search task into manageable segments that can be processed sequentially, improving accuracy without overwhelming system complexity
Solution Approach 2:
The first model performs preliminary action by selecting an initial set of candidate images based on text input before the second model conducts detailed matching. This preliminary filtering reduces the search space and improves the effectiveness of subsequent precise matching
2Measurement precision
If a single model is used for text-to-image search, then the system is simple, but the matching accuracy is insufficient
Solution Approach 1:
The single model is segmented into two distinct models with different functions: first model for broad candidate selection, second model for precise matching. This segmentation allows each model to specialize in its task, improving overall accuracy while maintaining reasonable system complexity
Solution Approach 2:
The first model acts as an intermediary between text input and the second model, transforming text queries into an initial image set that serves as input for the second model. This intermediary step bridges the gap between text and image domains, improving matching accuracy
Data Source
AI summary
A processing method includes obtaining character information, the character information being used to represent a search target, obtaining an image set, the image set including a plurality of images, and based on the character information, the image set, and an intelligent engine, obtaining an image search result including. Based on the character information and a first model in the intelligent engine, a first set is obtained. The first set includes a plurality of first images. Based on the first set, the image set, and a second model in the intelligent engine, a second set is obtained. The second set includes a plurality of second images, the second images is used as image search results, and the first model is different from the second model.