Operating algorithm of the automatic audio description system

Any information system operates in accordance with its internal algorithms. In the case of a system based on artificial intelligence, the development of a detailed algorithm becomes an even more important task: it is necessary to reduce the number of system errors in advance by calculating possible scenarios for its operation. So, for an automatic audio description system, it was decided to design an algorithm that will subsequently be used directly in the development of such a system. The article presents a detailed algorithm and description of the tables that this system accesses when going through the steps of the algorithm. It is assumed that by the last point of the algorithm, all the text descriptions necessary for dubbing the video sequence and the corresponding time intervals will be collected in the database.