Visual Similarity Grouping for Efficient Mobile Browsing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional approaches for finding visually similar items in large catalogs are inefficient and computationally expensive, often resulting in unwieldy presentations on smaller devices like mobiles, failing to accurately prune results based on user-selected visual attributes.
Innovation Solution
The method involves analyzing images to identify visual attributes such as color, pattern, and occasion, assigning these attributes to items of interest, and using a visual similarity score to group and rank similar items, allowing users to browse visually similar products in a more organized and efficient manner.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional image similarity search is used in large catalogs, then visual similarity can be determined, but the search process becomes slow and inefficient
Solution Approach 1:
The patent segments the image into multiple regions (e.g., foreground object, background, different parts of the item) and extracts visual attributes separately for each region. This allows the system to focus computation on relevant areas rather than processing the entire image, thereby maintaining visual similarity accuracy while improving search efficiency in large catalogs.
2Quantity of substance
If all visually similar items are presented in the results, then comprehensive results are provided, but the presentation becomes unwieldy on smaller devices
Solution Approach 1:
The patent extracts and emphasizes specific visual attributes (such as color, pattern, shape) from the similar items and uses these as organizing criteria. By grouping items based on shared visual attributes and presenting them in a structured manner, the system provides comprehensive results while making them manageable and easy to navigate on smaller device screens.
3Measurement precision
If manual attribute identification is used for each item, then accurate categorization is achieved, but the process becomes time-consuming and labor-intensive
Solution Approach 1:
The patent implements automatic visual attribute extraction through image processing algorithms that analyze the segmented image regions and identify attributes such as color, pattern, shape, and material. This self-service approach eliminates the need for manual attribute identification, maintaining high accuracy while dramatically reducing the time and labor required to categorize items in the catalog.
Data Source
AI summary
Various approaches discussed herein enable browsing groups of visually similar items to an item of interest, wherein the item of interest may be identified in a query image, for example. One or more visual attributes associated with the item of interest are identified, and the visually similar items matching at least one of the visual attributes are grouped together, wherein the group is ranked according to the visually similar items' overall visual similarity to the item of interest, for example by using a visual similarity score and/or metric.