TV production application based on Artificial Intelligence specific algorithms

The paper proposes an application that integrates an artificial intelligence model and is intended for the television production environment. This uses a pre-trained model, capable of quickly and accurately identifying relevant video files in a vast archive. This process involves the automatic analysis of metadata and audio-visual content to detect the information sought in each file, resulting in a list of files that match the selection criteria made with the help of implemented YOLO-type algorithms. The application is powered by an artificial intelligence model that can be customized and continuously created through additional training, using input data provided by users. Thus, an efficient and adaptable solution is created, which optimizes archive management and improves the production flow. The proposed application aims to eliminate the time spent identifying audio-video material necessary for editing.