Scene Intelligence Features

The Scene Intelligence Endpoint analyzes multimedia content to provide detailed insights into different aspects of a scene. The endpoint describes the scene setup, including spatial arrangements and visual elements, and identifies characters, both human and animal. It also detects branding, such as logos or product placements, and evaluates their visibility and impact. Finally, it processes text, identifying its location and meaning within the scene.

Features dictionary

Focal Points: The average focal points per creative, as well as the number of focal points per scene. Focal points refer to objects in the foreground that command attention.

Branding Area: % of visual area in a scene taken up by branding elements

Number of Branding Elements: # of branding elements in a scene

Time to Earliest Branding: The time, in seconds, to the first appearance of a branding element, like a logo or product placement.

Scene Description: A summary of the scene that includes its mood (e.g., cheerful, tense), any visible text transcription, and the overall tone of interactions or events.

Emotional Tone: The emotional tone conveyed by the most representative image of the scene in 1 word among: Contentment, Melancholy, Excitement, Anxiety, Serenity, Irritation.

Timing: Duration of a given scene (in seconds)