Video processing in audiovisual productions

At IT’S ON, we are developing an AI-powered Script Supervisor Assistant that aims to transform how audiovisual productions connect what’s written with what’s filmed.

Our goal is to build a system capable of analyzing all live feeds and playbacks from a set, and mapping them automatically to pre-production materials and scripts. This requires combining computer vision, video processing, and natural language understanding so that the AI can both “see” what happens in the footage and “read” what’s written in the script.

The expected outcome is an engine that can:

• Map video takes and playbacks to their corresponding script lines and scenes.

• Detect inconsistencies or continuity errors between the filmed material and the intended script (raccord analysis).

• Generate automated reports, such as script breakdowns or daily shooting parts, based on what was actually filmed.

We are currently looking for collaborators, researchers, and technical partners in the fields of:

• Computer vision and video analysis

• Natural language processing (NLP)

• Machine learning / AI model integration

• Film production workflows and metadata systems