Skeleton-Based Customer Interest Detection Without Product-Specific Rules
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies require manual generation of detection rules for each commodity product to identify customers highly effective in customer service, which is impractical due to the large number of products, making it difficult to effectively detect such customers.
Innovation Solution
A system that generates detection rules by analyzing past behavior and image data to associate customer motions with commodity product attributes and interest levels without relying on specific product types, using a combination of cameras, POS data, and an information processing apparatus to track and recognize skeletal motions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual detection rules are generated for each commodity product, then detection accuracy for customer interest is improved, but the complexity and time required for rule generation increases significantly
Solution Approach 1:
The patent segments the detection process into two parts: (1) generic skeleton-based motion detection that applies to all products, and (2) product-specific attribute extraction from image data. This segmentation allows the system to avoid creating complex rules for each product while maintaining detection accuracy through the combination of general motion patterns and specific product attributes.
Solution Approach 2:
The patent creates a universal detection framework using skeleton information that can be applied across all commodity products. The skeleton-based motion detection system serves multiple functions: detecting customer presence, tracking customer movement, identifying interest behaviors, and correlating with product attributes. This universal approach eliminates the need for product-specific rule generation while maintaining detection capability.
2Measurement precision
If detailed detection rules are created for each product, then customer interest detection precision is improved, but the time required for implementation increases
Solution Approach 1:
The patent performs preliminary action by pre-defining skeleton-based motion detection patterns and product attribute extraction methods that can be universally applied. The system pre-processes image data to extract product attributes (such as product ID, position, and type) before customer interaction analysis, so that when a customer exhibits interest behavior, the system can immediately correlate the skeleton motion with the pre-extracted product information without time-consuming rule matching.
3Extent of automation
If the system analyzes skeleton information to detect customer behavior, then automation level is improved, but the complexity of data processing increases
Solution Approach 1:
The patent extracts only the essential skeleton information (joint positions and movements) from full image data, discarding unnecessary visual details. This extraction approach enables automated detection of customer behaviors (such as approaching, viewing, or reaching for products) while keeping data processing complexity manageable by focusing only on the critical motion patterns needed for interest detection.
Data Source
AI summary
An information processing apparatus detects a person and a commodity product from image data. The information processing apparatus acquires, from the image data, a position of a skeleton of the person included in skeleton information on the detected person. The information processing apparatus specifies, based on the position of the skeleton of the person, a behavior of the person exhibiting with respect to the commodity product. The information processing apparatus specifies, based on the specified behavior of the person exhibiting with respect to the commodity product, a combination of an attribute of the commodity product and a degree of interest in the commodity product.