People-Group Cataloging via DBSCAN Clustering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional systems face difficulties in organizing photos into albums based on relationships between people, as they struggle to identify and manage multiple social circles, leading to cumbersome manual processes and inefficient automated methods that fail to recognize individuals' membership in multiple groups.
Innovation Solution
A customized density-based spatial clustering of applications with noise (DBSCAN) algorithm is used to analyze relationships between people in photos, determining optimal parameters based on location information and clustering photos based on similarity to identify people-groups, allowing for dynamic grouping and efficient album organization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If conventional systems manually organize photos into albums based on people, then photos can be organized into albums, but the process becomes cumbersome and time-consuming
Solution Approach 1:
The system automatically identifies people in photos, determines their relationships through clustering analysis, and organizes photos into albums without user intervention. The DBSCAN algorithm self-organizes the photo dataset by analyzing spatial relationships between people across multiple photos, eliminating the need for manual album creation while maintaining high accuracy in relationship identification.
2Extent of automation
If conventional systems automatically identify people in photos, then photos can be organized by individual people, but the systems fail to identify relationships between people and create albums based on social circles
Solution Approach 1:
The system transitions from one-dimensional individual photo organization to multi-dimensional relationship-based organization. By applying DBSCAN clustering to photo data, the system creates additional dimensional layers that represent social relationships, enabling photos to be organized not only by individual people but also by their relationships and social circles, thus preserving and utilizing relationship information that conventional systems lose.
3Adaptability or versatility
If conventional systems create albums for each identified person, then individual person albums can be created, but the systems cannot recognize that a person belongs to multiple social circles
Solution Approach 1:
The photo organization system achieves multi-functionality by enabling a single photo to belong to multiple albums simultaneously based on different social circles. The DBSCAN clustering algorithm identifies multiple relationship groups, and the system automatically creates multiple albums where a person's photos can be distributed across different social circle albums, making the system adaptable to complex social structures without requiring manual configuration of album hierarchies.
Data Source
AI summary
Methods and systems are provided for determining intelligent people-groups based on relationships between people. In embodiments, a photo dataset is be processed to represent photos of the photo dataset using vectors. These vectors include the importance of people in the photos. The photos are analyzed to determine similarity between the photos. Similarity is indicative of relationships between the photos of the photo dataset. The similarity is based on the people in the photos. The photos are clustered based on the similarity. In clustering the photos, clustering parameters determined from location information associated with the photos of the photo dataset are used.