What it does
This KPI uses advanced face recognition technology to detect faces in marketing assets in the first scene of a video asset.
How it works
The image is analyzed by a deep learning model trained to recognize human faces. For each image it returns the location of a face detected at or above a confidence level.
Why it matters
Humans tend to allocate attention to other people. By incorporating a human in the first scene, the ad has a higher chance to attract attention and create engagement.
How to achieve great results
Include human(s) in the opening scene to encourage audience engagement and captivate their attention from the start of the ad.
AI models used
Face detection models at aimpower are based on the latest deep learning frameworks used in scientific literature for the purpose of memorability modelling, including convolutional deep learning networks and vision transformers. The images used to collect human ground truth data originate from various sources, including social media platforms, large scale webscrapes (Common Crawl) and marketing materials (e.g. ads).
In total we leverage 1m+ images for pretraining, and 32K+ images with ~400,000 human annotated faces for fine-tuning our models on the face detection task. This training enables the model to accurately recognize and analyze faces across different scales, poses, expressions, occlusions, and lighting conditions. Evaluation of our models follows the scientific best practices. Accuracy versus human ground truth data is 99%.