How Technology Will Be Used to Control the Crowds During the World Cup

  • Publish date: Monday، 27 June 2022
How Technology Will Be Used to Control the Crowds During the World Cup

Crowd management will be one of the greatest problems Qatar will have to deal with by the end of the year. Researchers are currently looking for a solution.

New technology has been identified by research teams in Qatar to aid crowd control at the much-anticipated FIFA World Cup later this year.

To manage the anticipated 1.5 million people who are expected to swarm the Gulf nation for the event in November, the College of Engineering team at Qatar University (QU) suggested using cutting-edge technologies including surveillance drones, ICT, and AI.

The organizing committee for the FIFA World Cup Qatar 2022 in Qatar is primarily concerned with the security and safety of competitors, spectators, and other stakeholders.

Crowd control at the World Cup stadiums and their perimeters is crucial to ensure the safety and effectiveness of the activities due to the anticipated inherent opacity and density of the audience inside and outside the stadiums.

Google fires the engineer who claimed that an AI chatbot had become "sentient"

In collaboration with the Supreme Committee for Delivery and Legacy, the state-owned institution developed an intelligent crowd management and control system with multiple components for crowd counting, face recognition, and abnormal event detection (SC).

How will it help?

The method uses drone data to count crowds while extracting valuable features and estimating crowd density using dilated and scaled neural networks.

During the competition, the Football Supporters Crowd Dataset (FSC-Set), a brand-new dataset for crowd counting at sporting events, will also be unveiled. It includes 6000 manually categorized images of diverse scenes with large crowds gathered in or around the stadiums.

The research group has also focused on developing a face detection system that uses a multi-task convolutional neural network to consider faces in different positions. A cascade structure was specially employed to integrate a posture estimation approach and a face identification module.

The training data for the CNN-based posture estimate method was face captured from the left side, front, and right side.

To identify faces based on the estimated pose, three CNN architectures—VGG-16+PReLU left, VGG-16+PReLU front, and VGG-16+PReLU right—have also been implemented.

Meanwhile, a skin-based face segmentation method based on structure-texture decomposition and a color-invariant description has been given in order to remove extraneous face information (e.g., background content). In empirical tests on four well-known face recognition datasets, the proposed method outperformed related state-of-the-art methods.

Due to its dependability and affordability, abnormal event detection (AED) has recently grown in popularity thanks to drone-based video monitoring. Drones with cameras are typically capable of spotting hostile behavior among spectators during sporting events.

Thus, using a deep multiple instance rating systems and training video sequences with inadequate annotation, aberrant events are learned. In this manner, complete movies rather than specific scenes are annotated with instructional information.

Follow us on our Whatsapp channel for latest news