3D Face Pose Estimation for Retail Viewership Measurement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for measuring viewership of displayed objects in retail environments fail to accurately distinguish between passers-by and actual viewers, leading to inaccurate assessments of advertising effectiveness, as they rely on assumptions about head orientation or use cumbersome infrared-based systems that are impractical and costly.
Innovation Solution
A system utilizing computer vision technologies, including face detection, tracking, and 3-dimensional face pose estimation, to count actual viewers and measure the duration of their engagement with displayed objects, differentiating between viewers and passers-by, and providing data on impression levels and media effectiveness without the need for manual input or cumbersome devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If infrared-based systems are used to measure viewership, then measurement capability is provided, but device complexity and cost increase significantly
Solution Approach 1:
The patent replaces complex infrared-based measurement systems with a computer vision system using standard cameras and image processing algorithms. Instead of relying on sophisticated infrared sensors and hardware, the invention uses software-based face detection, tracking, and pose estimation to determine viewership, thereby reducing device complexity while maintaining measurement capability
Solution Approach 2:
The patent creates a virtual representation of the physical measurement system through software simulation. By capturing images with standard cameras and processing them through computer vision algorithms, the system creates a digital model of viewer behavior without requiring complex physical infrared sensors, thus simplifying the hardware while preserving measurement function
2Productivity
If assumptions about head orientation are used to measure viewership, then measurement can be performed, but measurement precision deteriorates
Solution Approach 1:
The patent implements continuous feedback through real-time image capture and processing. The system captures multiple images, tracks face orientation changes, and provides ongoing feedback about actual viewer engagement. This feedback mechanism allows the system to distinguish between passers-by and actual viewers by monitoring changes in head orientation and gaze direction, thereby improving measurement precision while maintaining efficiency
Solution Approach 2:
The patent performs preliminary actions by capturing images at multiple time points and analyzing head orientation changes before finalizing viewership determination. By pre-processing the image data and tracking facial pose evolution over time, the system can accurately identify viewers who maintain engagement with the displayed object, improving measurement precision without sacrificing productivity
3Measurement precision
If manual input methods are used to measure viewership, then measurement accuracy can be maintained, but ease of operation deteriorates
Solution Approach 1:
The patent implements self-service by enabling the system to automatically capture, process, and analyze viewer behavior without manual intervention. The computer vision system autonomously performs image capture, face detection, tracking, and pose estimation, eliminating the need for manual data collection or operator involvement. This automation maintains measurement precision while dramatically improving ease of operation
Solution Approach 2:
The patent replaces manual measurement methods with automated computer vision technology. Instead of requiring human operators to observe and record viewer behavior, the system uses digital image processing and artificial intelligence algorithms to automatically determine viewership, thereby maintaining accuracy while eliminating operational complexity and improving ease of use
Data Source
AI summary
The present invention is a method and system for measuring viewership of people for a displayed object. The displayed object can be specific in-store marketing elements, such as static signage, POP displays, and other forms of digital media, including retail TV networks and kiosks. In the present invention, the viewership comprises impression level, impression count of the viewers, such as how many people actually viewed said displayed object, average length of impression, distribution of impressions by time of day, and rating of media effectiveness based on audience response. The viewership of people is performed automatically based on the 3-dimensional face pose estimation of the people, using a plurality of means for capturing images and a plurality of computer vision technologies on the captured visual information. The present invention distinguishes viewers from passers-by among the plurality of persons in the vicinity of the displayed object, by counting the number of viewers who actually viewed the displayed object vs. passers-by who may appear in the vicinity of the displayed object but do not actually view the displayed object, using the 3-dimensional face pose estimation and a novel usage of a plurality of computer vision technologies.