Brand Identification from Football (Soccer) Match Videos
Media & EntertainmentBusiness Impacts
Reduced resource and auditor cost
Decreased manual efforts for reviewing ad clips
Generate insights on frequency of ad occurrence and perform advanced analytics
Customer Key Facts
- Country : United Kingdom
- Industry : Advertising Technology
Problem Context
The client is a British company that provides LED technologies, live event production, and creative content services to deliver excellent audience experiences to their customers. The client wanted to identify brands from advertisements playing on LED screens in stadiums during football matches. Their customers have two auditors that manually review the video clips, and the client sought to save their customers’ efforts and costs by freeing up one auditor with an automated solution.
Challenges
- Time consuming process of manually reviewing the match clips for ad details.
- High error rates due to manual review of match clips.
Technologies Used
Amazon Rekognition
Amazon SageMaker
AWS Lambda
Amazon S3
Amazon CloudWatch
Identifying brands from match videos and enabling customers to increase their brand visibility.
Solution
Quantiphi is developing a computer vision-based solution that will assist the client in detecting and identifying brands in advertisements. Through text and brand logo detection, the solution will capture advertisements that are playing in the background on the LED screens during football matches.
In addition, the solution will provide a summary report at the end that specifies the brand, number of times the brand appeared on the screen, and the timestamps at which it occurred.
Results
- Enable cost savings and reduce manual efforts.
- Improve brand visibility for customers by using captured data to better understand brand exposure.
