IAB Content Taxonomy

Overview

The IAB Content Taxonomy 3.0 is the standard for the entire advertising ecosystem to segment and categorize content. Hive's APIs detect and return results for supported classes from this taxonomy. A key use case for this API is content-based ad targeting: enabling media companies to sell more granular contextually-targeted ad spots. The classes below refer to the image-based API; support for text and audio will be available separately.

For users specifically interested in the Brand Safety and Suitability classes mapped to the GARM taxonomy, another API is separately available here.

Supported IAB Classes

Hive currently addresses more than 130 IAB classes that are feasible to identify with visual signals; a comprehensive audio and text model mapping to all IAB classes from additional audible / textual signals will be released in the near future.

Note on Unique IDs: IAB provides a Unique ID for each of its classes. For convenience, these IDs are appended to the beginning of the each Hive Response ID.

*Parent ID is not included within our supported classes. For the full IAB Content Taxonomy see here.

Expected Output

For each frame delivered, the API will return a response for each class. The majority of classes will return either a 0 (negative) or 1 (positive). A small subset of classes (GARM classes only) will return a value from 0 to 1, indicating the confidence level of the classification.