Pre-Publication V2. Models trained on film composition that are intended designed to can predict cinematic framing by automatically layering elements including, but not limited to, characters, objects, backgrounds, and lighting. For example, it interprets narrative scripts or storyboards to predict how cameras, environments, and characters should be organized to maximize emotional impact and visual storytelling. (OR OPTION 2: delete this term)
Deliberation Summary:
(1) Clarify whether the term refers to a single capability or several distinct workflows presented together, and whether it is in common use; considered deleting. (2) Language clarifications suggested as systems compute statistical relationships rather than respond to emotion; example was not apt. (3) Language edits to describe what the models are designed to do rather than unqualified conclusions.
Pre-Publication V1. Models that can predict cinematic framing by automatically layering elements including, but not limited to, characters, objects, backgrounds, and lighting. For example, it interprets narrative scripts or storyboards to predict how cameras, environments, and characters should be organized to maximize emotional impact and visual storytelling.