Context-Aware Search Type Selection for Mobile Visual Queries
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image-based visual search technologies perform poorly when provided with certain types of image inputs, such as a sweater, leading to imprecise or inapplicable search results due to the lack of an optimal search tool or technology tailored to the user's context.
Innovation Solution
A method and system that determine a geographic location of a mobile computing device and obtain associated search types, then output a graphical user interface indicating these search types, allowing users to select and perform relevant searches, such as image-based visual searches, based on their location and context.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single search tool is used for all image inputs, then the device complexity is reduced, but the search precision deteriorates for certain image classes
Solution Approach 1:
The system dynamically selects search tools based on image classification results. Instead of using a fixed search tool, the system adapts the search approach according to the identified image class, allowing optimal search precision for each image type while managing complexity through conditional logic
Solution Approach 2:
The system changes the search parameter (search tool type) based on the image class identified. Different search tools are selected for different image classes (e.g., barcode search for product images, visual search for fashion items), optimizing search precision for each category
2Measurement precision
If context-specific search capabilities are provided, then the search precision is improved, but the device complexity increases
Solution Approach 1:
The system segments the search functionality into multiple specialized search tools, each optimized for specific image classes. The image classification module divides images into categories, and corresponding search tools are selected for each segment, providing context-specific precision without requiring all tools to be active simultaneously
Solution Approach 2:
The system implements a universal search interface that can perform multiple search types through a single unified entry point. The search tool selection is automated based on image classification, making the system multi-functional while maintaining a simple user interface that doesn't expose the underlying complexity
3Ease of operation
If image-based visual search is used for all search types, then the ease of operation is improved, but the search precision deteriorates for non-visualizable objects
Solution Approach 1:
The system automatically changes the search parameter (search tool type) based on the image class identified. For images containing barcodes, text, or structured data, alternative search methods are selected instead of visual search, maintaining both ease of operation through automation and search precision by matching the search method to the content type
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Example techniques and systems may obtain one or more search types associated with a geographic location of a computing device. In one example, a technique may include determining, by a mobile computing device, a geographic location of the mobile computing device and obtaining one or more search types associated with the geographic location. Responsive to obtaining the one or more search types, the technique may also include outputting, by the mobile computing device and for display, a graphical user interface comprising at least an indication of the one or more search types associated with the geographic location.