Pose-Aware Image Retrieval Query Generation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image retrieval methods face challenges in accurately and efficiently retrieving images based on pose information, as they often rely on manual metadata application and lack effective utilization of pose information for query generation, leading to suboptimal retrieval accuracy and efficiency.
Innovation Solution
An image retrieving apparatus and method that includes a pose estimating unit, features extracting unit, image database, query generating unit, and image retrieving unit, which recognize and extract pose information from input images, generate retrieval queries, and retrieve images with similar poses, thereby improving retrieval accuracy and efficiency by incorporating pose information into the query process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If manual metadata application is used for image retrieval, then retrieval can be performed on large-scale image data, but retrieval accuracy and efficiency deteriorate due to the enormous data volume and manual effort required
Solution Approach 1:
The patent replaces manual metadata application with automated pose information extraction using pose estimating units and feature extracting units. This substitution of mechanical manual operations with automated computational systems enables efficient processing of large-scale image data without proportional increases in manual effort, thereby maintaining high retrieval efficiency while handling enormous data volumes
Solution Approach 2:
The system enables self-service by automatically extracting pose information and generating retrieval queries without requiring manual metadata application. The pose estimating unit and feature extracting unit autonomously process images to generate retrieval-ready data structures, allowing the system to serve itself in preparing data for retrieval operations
2Measurement precision
If pose information is extracted and used for query generation, then retrieval accuracy improves, but device complexity increases due to additional processing units
Solution Approach 1:
The patent implements multi-functionality by designing the pose estimating unit and feature extracting unit to serve multiple purposes: they not only extract pose information for retrieval queries but also generate structured data that can be used for various types of image searches. This universal approach allows a single system structure to handle diverse retrieval needs, reducing overall system complexity while maintaining high retrieval accuracy
Solution Approach 2:
The system performs preliminary action by pre-extracting and storing pose information in the image database during image ingestion. This advance preparation of pose features eliminates the need for complex real-time processing during retrieval operations, thereby improving accuracy without proportionally increasing operational complexity
3Reliability
If pose information is utilized in retrieval queries, then user intention alignment improves, but loss of time occurs in pose estimation and feature extraction processes
Solution Approach 1:
The patent applies preliminary action by extracting and storing pose information during the image ingestion phase rather than during retrieval operations. The pose estimating unit and feature extracting unit prepare pose features in advance, storing them in the image database associated with each image. This advance preparation ensures user intention alignment is maintained while avoiding time loss during actual retrieval, as the pose data is already processed and ready for immediate query matching
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An image retrieving apparatus includes a pose estimating unit which recognizes pose information of a retrieval target including a plurality of feature points from an input image, a features extracting unit which extracts features from the pose information and the input image, an image database which accumulates the features in association with the input image, a query generating unit which generates a retrieval query from pose information specified by a user, and an image retrieving unit which retrieves images including similar poses according to the retrieval query from the image database.