Query Change System for Image Search Contradiction Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image search systems may output images that contradict search target conditions when query images do not fully satisfy the required criteria, leading to irrelevant search results.
Innovation Solution
A query change system that includes a processor to identify and correct contradicting parts in query images based on search target conditions, generating a revised query image that aligns with the intended search criteria, thereby ensuring more accurate search results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a search is performed using only features extracted from a query image, then the search process is simple and fast, but images matching features contradicting the conditions related to search targets may be outputted, reducing search accuracy
Solution Approach 1:
The system performs preliminary action by detecting contradicting parts in the query image before the search is executed. The contradiction detection unit identifies features that contradict search target conditions, and the correction unit modifies these features in advance, ensuring that the subsequent search operates on corrected features that align with the intended search criteria, thus preventing inaccurate results without significantly delaying the search process
Solution Approach 2:
The system implements feedback by using the contradiction detection unit to analyze the query image features against search target conditions, identify contradictions, and provide this information to the correction unit. The correction unit then adjusts the features based on this feedback, creating a closed-loop system that continuously refines the search features to eliminate contradictions while maintaining search efficiency
2Measurement precision
If query images are corrected to match search target conditions, then search accuracy is improved, but the device complexity increases due to contradiction detection and correction mechanisms
Solution Approach 1:
The system applies segmentation by dividing the contradiction detection and correction process into distinct functional units: a contradiction detection unit that identifies contradicting features, and a correction unit that modifies them. This modular segmentation allows each unit to perform its specific function independently, making the overall system more manageable and maintainable while achieving accurate search results through coordinated operation of these specialized components
Solution Approach 2:
The correction unit acts as an intermediary between the query image features and the search process. It receives features from the query image, detects contradictions against search target conditions, and produces corrected features that are then used for searching. This intermediary role isolates the complexity of contradiction handling from both the query input and the search execution, simplifying the overall system architecture while maintaining high search accuracy
Data Source
AI summary
A query change system includes: a processor configured to correct, in a case where a first query image inputted by a user includes a contradicting part that contradicts a first condition related to a search target, the contradicting part of the first query image in accordance with the first condition to generate a second query image.