Raksul
Background
With the rise of internet shopping and multi-channel marketing, understanding the impact of TV commercials is crucial for businesses like Raksul Japan. However, TV ads are aired during broad, unspecific time slots throughout the week or month, making it difficult to assess the exact timing and impact of individual ad campaigns. To address this, Raksul needed a platform capable of identifying and isolating TV commercials from recorded television streams as a first step toward analyzing the effectiveness of ad campaigns.
Challenge
TV commercials appear within varying timeframes on different channels, and there was no reliable method to automatically detect and separate these ads from hours of recorded TV streams. Raksul required a solution that could precisely detect ad sequences without relying on traditional blank screen transitions, ensuring that the detection would work across a wide range of TV channels and content formats.
Solution
Vidatt developed an AI-driven algorithm to detect and isolate individual TV commercials with high precision. Key features of the solution included:
Ad Sequence Detection
Using advanced AI techniques, Vidatt designed an algorithm capable of detecting ad sequences in recorded TV streams without depending on conventional indicators like blank screen transitions
Ad Separation
The algorithm could accurately separate each specific ad within a commercial break, providing individual ad sequences for further analysis
Scalable Testing
To ensure robustness, the algorithm was tested over 500 hours of broadcast content across various TV channels. This large-scale testing validated the solution’s accuracy and effectiveness
Our Contribution
Vidatt’s contributed to form a high performing algorithms enabled Raksul to optimize its ad campaign analysis, providing valuable insights into the success of its TV advertising strategies and paving the way for more data-driven marketing decisions.
Outcome
Vidatt’s algorithm achieved over 96% accuracy in detecting and separating TV commercials from recorded streams, significantly streamlining the process of ad impact analysis for Raksul. With this capability, Raksul could further analyze the detected ads to assess their effectiveness, including identifying the products and personalities featured in the commercials