AINAR
Ad Breaks & Contextual Ads
Revolutionizing Advertising with Scene-Level Precision
We developed Ad Breaks and Contextual Ads to help AVOD and FAST services increase their advertising revenue while providing a seamless viewing experience for their audiences. By utilizing the latest advances in computer vision and machine learning, AINAR is able to detect natural advertisement breaks in video assets – accurately and efficiently. With AINAR ad breaks, users don’t feel as if commercials are randomly shoved in during a gripping scene or in the middle of an important dialogue, as ads are placed at the natural cue points in the story. AINAR Ad-Breaks divide breaks into three ranks. Primary breaks, secondary, and tertiary breaks allow content owners to prioritize and tier their advertisements.
AINAR Contextual Ads accurately detect faces, actions, scenes, emotions, and moods to serve ads in line with the content you are consuming. Through contextual ad breaks and detailed scene analysis, AINAR enables hyper-targeting, leading to an increase in impressions and conversion. AINAR Contextual ADs “watch” a movie as a human would and describe what is going on in the scene leading up to the ad break. The scene data includes +3000 keywords and +700 mood tags. This helps advertisers keep emotions at the heart of their campaigns and ensure that ads are placed in the right emotional context.
Automated mapping into IAB taxonomy 3.0 for programmatic targeting.
The Ranking Structure, how it works
Four parallel networks work in harmony to predict and rank ad breaks into three distinct categories:
- Network 1: Black frame detector
- Network 2: Deep learning network that looks at the flow of the story
- Network 3: Scene boundary detector
- Network 4: Shot-similarity algorithm
Using these 4 networks, we are able to create a ranking system:
- Rank 1 demands a full consensus from all networks, signifying the highest quality ad.
- Rank 2 necessitates agreement from at least three networks
- Rank 3, necessitates agreement from at least two networks
- Rank 4, Gap Filler, includes ad breaks only predicted by the shot similarity network, and is used to fill gaps when higher-ranked ad breaks are insufficient.
We also do a post-processing filter for speech to make sure we are not cutting off in an important discussion or narration that travels across scenes.
Ad Breaks
What´s included
ad breaks
Ad Breaks creates a seamless advertising experience for viewers as ads are placed at natural cue points in the story.
RANKING OF AD BREAKS
AINAR Ad-Breaks divide breaks into three ranks. Primary breaks, secondary, and tertiary breaks allow content owners to prioritize and tier their advertisements effectively.
Contextual Ads
What´s included
contextual scene metadata
Contextual scene metadata that describes what was going on in the scene leading up to the ad break. Includes +3000 keywords and +700 mood tags. Automated mapping into IAB taxonomy 3.0 for programmatic targeting.
RANKING OF AD BREAKS
AINAR Ad-Breaks divide breaks into three ranks. Primary breaks, secondary, and tertiary breaks allow content owners to prioritize and tier their advertisements effectively.
We help TV-Operators, IPTV operators, cable operators, OTT platforms, and broadcasters create enhanced viewing experiences, release the potential of all their content and provide content that aligns with each viewer’s taste.
Contact
Västmannagatan 4
111 24 Stockholm, Sweden
