Deep Learning content-based search in media files

The rapid growth of digital multimedia content has created a new demand for technologies that not only process but also sort and locate images and videos amidst the vast amounts of content. Traditional text-based search engines often struggle to retrieve specific visual content based on visual features alone, which has spurred the development of computer vision technologies for content-based image and video retrieval. This work presents the design, implementation, and evaluation of a user-friendly application, "Search by visual content," that leverages artificial intelligence (AI) and, more particularly, the deep learning model of YOLOv8. Our application allows users to search multimedia sources such as images and videos not by their names, types, and metadata, but by using similar images or videos as input. It also offers the feature of searching multimedia content containing a certain object inputted in the search box. Our easy-to-use application delivers prompt and accurate search results, and it can function across multiple operating systems as it was made using Python language.