Metrics for Anonymous Video Analytics on Digital Signage
Anonymous Video Analytics (AVA) is a passive and automated audience measurement technology designed for digital signage networks that can be used to provide digital signage operators with quantitative viewership information. Although AVA promises to bring accountability into the digital signage industry, there are still no industry standard metrics that can be used to assess the accuracy of any particular AVA system.
In this paper we will describe a methodology and present a set of accuracy and error metrics that we believe will help to characterize the various parameters in most AVA systems, such as viewer counts, dwell times, and demographic (gender/age) classification accuracy.
The basic question this paper attempts to address is whether you can trust the anonymous video analytic data which you have collected with a particular system. By demonstrating how these metrics can be used on an actual AVA system, we hope to move the industry towards a standard set of metrics that may help advertisers and digital signage operators make a more informed decision when assessing the various video analytics solutions available on the market. Our metrics can also be used by AVA solution providers to determine the strengths and weaknesses of their systems in order to improve the quality of their measurement technology.
Read the full Metrics for Anonymous Video Analytics on Digital Signage White Paper: http://www.intel.com/content/www/us/en/retail/digital-signage-ava-metrics-paper.html