How does Magellan AI determine if an advertisement is host-read?
To flag detected advertisements as either host-read or produced at scale, Magellan AI uses a heuristic-based rule analyzing all voices detected in a podcast episode. Voices matched in an ad read found elsewhere in the podcast show content are flagged as host-read. If a voice does not match any other voices detected in the show content, the ad is flagged as produced.
Please note this is a beta feature and Magellan AI users are able to flag incorrectly identified advertisements.
Updated almost 2 years ago