Systema for Automatic Generation of Adaptative Audiodescription
Accessibility, Adapted Audio description, Deep Learning, Video Description
Audio description is an accessibility resource designed to make visual information accessible to blind or low-vision people. To increase the range of audio description in digital video applications, we propose a system for automatic generation of audio description for films. The system can use as source of information about the film the original script and the video itself. As proof of concept, we developed a prototype that generates audio description scripts adapted to the preference of users with visual impairments based on actions taken from the script and objects recognized in the video. The experiments contemplated the application of the solution in fiction films and surveillance videos. The partial results show that the proposed solution has the potential to generate descriptions of the most important events of the videos tested. The correct identification of the scene object shows that the approach is in the right direction for the automatic audio description only based on video