AR Product Recommendation via Style and Color Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional techniques for creating targeted digital content are often inaccurate due to user reluctance to provide information and reliance on human-generated metadata, which is time-consuming and expensive, and fail to account for physical surroundings not reflected in digital footprints.
Innovation Solution
An augmented reality (AR) system that identifies objects in a user's physical environment through a camera feed, using style similarity and color compatibility algorithms to generate product recommendations that match the real-world surroundings, providing accurate and flexible product suggestions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional techniques use user-provided information and web-browsing data for targeted content, then personalization is achieved, but accuracy deteriorates due to user reluctance to provide information and intentional false information
Solution Approach 1:
The patent introduces an intermediary mechanism (style extraction algorithm) that indirectly infers user preferences through objective analysis of physical environment styles rather than directly relying on user-provided information. This mediator bridges the gap between user needs and accurate recommendations without requiring trustworthy user input.
Solution Approach 2:
The patent replaces the mechanical system of collecting and trusting user-provided data with an automated visual analysis system using computer vision and style extraction algorithms. This substitution eliminates reliance on user honesty while maintaining personalization accuracy through objective environmental assessment.
2Productivity
If conventional techniques rely on human-generated metadata for product recommendations, then product categorization is achieved, but productivity deteriorates due to time-consuming and expensive manual tagging processes
Solution Approach 1:
The patent implements self-service by enabling the system to automatically extract style characteristics directly from images of physical environments without requiring external human annotation. The system serves itself by using computer vision algorithms to perform what previously required human metadata creators, eliminating time loss while maintaining recommendation quality.
Solution Approach 2:
The patent substitutes the manual human tagging mechanism with an automated computer vision-based style extraction system. This replacement eliminates the time-consuming and expensive human labor while achieving the same or better categorization accuracy through algorithmic analysis of visual features.
3Adaptability or versatility
If conventional techniques use metadata-based product recommendations, then product suggestions are generated, but adaptability deteriorates when metadata is unavailable or inaccurate
Solution Approach 1:
The patent inverts the conventional approach by not starting from product metadata but rather starting from the visual analysis of the physical environment. Instead of matching products to predefined metadata categories, the system extracts style characteristics directly from images and matches products to these extracted styles, enabling adaptability when metadata is unavailable.
Solution Approach 2:
The patent introduces style extraction from visual data as an intermediary that bridges the gap between physical environment and product recommendations. This intermediary allows the system to adapt to any environment regardless of metadata availability, maintaining reliability through objective visual analysis rather than dependent on potentially missing or inaccurate metadata.
4Measurement precision
If conventional techniques focus on digital footprint data, then user profiling is achieved, but measurement precision deteriorates by failing to capture physical surroundings
Solution Approach 1:
The patent transitions from the digital dimension (web-browsing data, user profiles) to the physical dimension by capturing and analyzing images of the user's physical environment. This dimensionality change enables measurement of physical surroundings that were previously inaccessible, improving accuracy of environmental context while recovering lost information about the user's real-world space.
Solution Approach 2:
The patent substitutes the digital data collection mechanism with a visual capture mechanism using camera feeds and image analysis. This replacement enables direct observation and measurement of physical surroundings, achieving precision in capturing environmental context that digital footprints alone cannot provide.
Data Source
AI summary
The present disclosure relates to systems, methods, and non-transitory computer readable media for generating augmented reality representations of recommended products based on style similarity with real-world surroundings. For example, the disclosed systems can identify a real-world object within a camera feed and can utilize a 2D-3D alignment algorithm to identify a three-dimensional model that matches the real-world object. In addition, the disclosed systems can utilize a style similarity algorithm to generate style similarity scores for products in relation to the identified three-dimensional model. The disclosed systems can also utilize a color compatibility algorithm to generate color compatibility scores for products, and the systems can determine overall scores for products based on a combination of style similarity scores and color compatibility scores. The disclosed systems can further generate AR representations of recommended products based on the overall scores.