Image Browsing System Using Dynamic Subset Generation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image classification and retrieval methods rely heavily on human input, which is tedious and may not match the user's taste or context, and objective analysis methods do not necessarily align with user preferences, especially for large collections or unannotated images.
Innovation Solution
A method for browsing digital images that involves user interaction to select images, determining subsets based on common categories, and displaying additional images with strength values that reflect user interest, using a combination of image categories and user data to automatically select and display related images, allowing for dynamic category creation and updating based on user behavior.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If classification is based on textual annotation or metadata labels requiring user explicit input, then the classification can be tailored to user taste, but the task becomes fastidious and many images remain unlabelled
Solution Approach 1:
The system automatically generates classifications and subsets by analyzing user browsing behavior patterns, allowing the system to serve itself by inferring user preferences from observed interactions rather than requiring explicit user input for each classification task
Solution Approach 2:
The system continuously monitors and analyzes user interaction patterns with images and subsets, using this feedback to dynamically refine and update classifications and subset recommendations, creating a closed-loop system that adapts to user taste over time
2Extent of automation
If classification is based on objective image content analysis using preset algorithms, then the process is automated and scalable to large collections, but the results do not necessarily match user preferences or changing context
Solution Approach 1:
The system transitions from static preset algorithms to dynamic classification that adapts in real-time based on observed user behavior patterns, allowing classifications to evolve and change as user preferences and context change
Solution Approach 2:
The system modifies classification parameters and subset criteria based on analyzed user interaction data, adjusting the weights and thresholds of classification algorithms to better align with individual user preferences rather than using fixed preset parameters
3Loss of information
If all images are displayed at once for user selection, then the user has complete information, but the interface becomes overwhelming and navigation becomes difficult
Solution Approach 1:
The system divides the large image collection into multiple organized subsets based on analyzed user preferences and behavior patterns, presenting images in manageable groups rather than overwhelming the user with all images simultaneously
Solution Approach 2:
The system adds organizational dimensions to image presentation by creating hierarchical subsets and categories based on user behavior analysis, allowing users to navigate through multiple levels of organization rather than facing a flat, unstructured list of all images
Data Source
AI summary
A method for browsing a collection of digital images on a soft-copy display comprising: receiving a collection of digital images; interactively user selecting a digital image using a user interface; determining a plurality of subsets of the digital images, wherein each subset shares a common category with the selected digital image; and displaying the subsets of digital images on the soft-copy display, together with the selected digital image.