CASE STUDY | UEFA Super Cup Final 2020

Keeping Sports Fans Safe

Puskas Arena


Location: Budapest, Hungary
Deployment Type: High-attendance sporting event

Features: Mask Detection
Use Cases: COVID-19 transmission prevention, anonymous video analytics

Summary: G2K’s Parsifal IoT platform helped 15,000 fans safely attend a major football event — the UEFA Super Cup Final — during a pandemic, with the help of SAFR’s anonymous mask detection.

The Challenge

The impact of COVID-19 is undeniable as communities all over the world implement health and safety measures to reduce virus transmission. After social distancing guidelines prevented spectator attendance at sporting events for much of the year, the Union of European Football Association (UEFA) decided to test the partial return of spectators at the 2020 UEFA Super Cup Final in Budapest, Hungary. In order to ensure the health and safety of spectators, staff, media, and players, a set of guidelines was determined in accordance with local regulations — capping stadium capacity at 30%, social distancing, temperature checks, and mandatory mask wearing.

These guidelines required new monitoring and enforcement at Puskas Arena including body temperature scans and mask compliance screening. It was also necessary for this process to be contactless and automated to prevent formation of queues and crowds at entry points and to avoid further personnel costs.

The SAFR solution

To address these challenges, UEFA partnered with G2K Group, a German tech company with global operations, to deploy its IoT platform, Parsifal, on 21 cameras across 7 entry points. Parsifal’s AI video analytics provided a new level of situational awareness, scanning more than 15,000 arriving spectators as well as staff and media. Two key features for this use case — anonymous face mask detection and demographic analysis — were provided by SAFR from RealNetworks. A G2K partner, SAFR’s core computer vision algorithms are deployed as part of the Parsifal platform helping provide actionable, real-time video analytics. 

With a 95% detection rate and ability to distinguish masks from other occlusions, SAFR can reliably detect whether a face appearing on camera is properly masked. With high accuracy for both surgical style and patterned fabric face masks, SAFR mask detection is designed for challenging real-world conditions. 

The deployment of Parsifal at the stadium entry points produced the desired outcome, successfully detecting all persons attempting to enter the event without a mask.

Andreas Schaer, head of Venue Operations at UEFA, was pleased with the outcome of the test, noting that the technology was, “easy to use and performed very well during the ingress procedures on match day.” Parsifal and SAFR allowed the UEFA to deliver a safe and secure major football event for the attending public, even during a global pandemic.

“The use of SAFR’s highly accurate mask detection algorithm strengthened the overall Parsifal solution, making it better able to serve customers dealing with the realities of a global pandemic. G2K is looking forward to many more projects and further strengthening its prosperous partnership with SAFR.”

Georg Rennenkampff, VP for Partnership Management at G2K

SAFR gets you safely back to work and fun

SAFR’s highly accurate face recognition platform is designed for high performance and rapid processing enabling it to detect and match millions of faces, masked and unmasked, in real time. With widespread use of PPE face masks it’s critical that SAFR is able to maintain high accuracy rates for face recognition even when faces are partially covered.

Mask Detection

95% mask detection rate

Recognition of Masked Faces

98.85% face recognition accuracy for masked faces


Occupancy counting & mask detection dashboard

Contactless Secure Access

Touchless access control for masked faces

Request a free demo or get in contact with one of our sales or technical representatives to learn more.