Following internal and external testing, Quividi is ready to introduce a key new features to its Audience Measurement Platform: live Footfall.
Footfall is a real-time traffic counter based on silhouettes detection, using a new class of neural networks. Much like the historical OTS (Opportunities To See), it strives to estimate the number of people passing in front of the camera, irrespective of where they look, which is also called "Impressions". Unlike OTS, however, it works in real time, and without minimal operating conditions.
Quividi's footfall detection range goes up to 40 meters with standard cameras (or 25 meters with larger field of view cameras), even in low video resolutions, with a strong accuracy (95% and above).
With Footfall counting, you will now be able to:
- measure traffic of large video walls and external locations
- get real-time impressions per content
- get a high-fidelity currency for programmatic purposes.
How to deploy Vidireports Footfall on your network
Footfall is a free upgrade to our Core and Pro offers, and will be released alongside Vidireports 7.5. However, it will not be active by default. In order to access this feature, you'll want to follow those simple steps:
- Use the Upgrade Manager tool to upgrade your network to Vidireports 7.5 - As usual, we recommend to upgrade a few locations, then deploy it on the whole network.
- Send an email to email@example.com - Footfall requires additional processing, and as such we need to verify your performances before deploying this feature. Note that Footfall's CPU requirement are irrespective of the video resolution and number of pedestrians.
For more information on Footfall CPU requirements, please visit this page.
What impact will Footfall have on your data?
Footfall will not change your existing dashboards, but simply replace the existing OTS numbers with the silhouettes detected.
Our testing shows that in most configuration, footfall will detect more traffic than the current OTS. Footfall also performs better than OTS in outdoor locations, in locations where the traffic doesn't face the camera, or where the camera is too far to detect faces reliably.
An increase of OTS detection will not be matched by an increase of Watchers - This will result in a lower Conversion Ratio.
If you have any questions about Footfall, please send an email to firstname.lastname@example.org.