Free-Form Visual Search Feature Selection for Intent-Relevant Results
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing visual search query systems provide results that are limited to holistic visual characteristics of an image, failing to reflect the user's true search intent, especially when focusing on specific visual features within the image.
Innovation Solution
A computer visual search system that allows for free-form user selection of visual features, enabling the construction of visual search queries based on specific sub-portions of an image, combining results from both the overall object and the selected features to provide more personalized search results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If visual search systems focus exclusively on holistic visual characteristics of query images, then search results are easier to generate, but the results fail to reflect the user's true search intent regarding specific visual features
Solution Approach 1:
The patent segments the visual search process into two distinct components: holistic image search and specific visual feature search. The system divides the query processing by separating the overall image characteristics from user-selected specific features, allowing independent processing and combination of results. This segmentation enables precise targeting of user intent while maintaining system manageability through modular architecture.
Solution Approach 2:
The patent adds a new dimension to visual search by incorporating user-selected specific visual features as an additional search criterion beyond holistic image matching. This creates a multi-dimensional search space where results are filtered and ranked based on both overall image similarity and specific feature matching, thereby improving result relevance without overwhelming system complexity through progressive dimension addition.
2Measurement precision
If visual search systems allow free-form user selection of specific visual features, then search result relevance improves, but user interface complexity increases
Solution Approach 1:
The system performs preliminary actions by automatically detecting and extracting potential visual features from the query image before user selection. This pre-processing step prepares the image data and identifies candidate features, reducing the cognitive load on users and simplifying the interaction process while maintaining the ability to perform precise feature-based search when users select their interests.
Solution Approach 2:
The patent introduces an intermediary layer between the user and the complex visual feature selection process. This intermediary automatically processes user input, maps selections to relevant visual features, and manages the complexity of feature extraction and matching algorithms, thereby preserving search precision while shielding users from underlying system complexity.
3Adaptability or versatility
If visual search systems process both holistic image characteristics and specific visual features, then search intelligence improves, but processing time increases
Solution Approach 1:
The system implements partial action by allowing users to select only the specific visual features relevant to their search intent, rather than processing all possible features. This selective approach reduces processing time by focusing computational resources on user-interesting features while still performing holistic image matching, achieving a balance between search intelligence and processing efficiency through user-guided feature selection.
Data Source
AI summary
A user can submit a visual query that includes one or more images with user free-form selected visual features of interest. Various processing techniques such as optical character recognition (OCR) techniques can be used to recognize text (e.g., in the image, surrounding image(s), etc.) and/or various object detection techniques (e.g., machine-learned object detection models, etc.) may be used to detect objects and particular visual features of objects (e.g., dress, sleeves, color, pattern, etc.) within or related to the visual query. Content related to the detected text or object(s) in combination with the user free-form selected visual feature of interest can be identified and potentially provided to a user as search results. As such, aspects of the present disclosure enable the visual search system to more intelligently process a visual query to provide improved search results and content feeds, including search results which are personalized to account for user search intent.