Motion Picture Identification using Audio Processing and Natural Language Processing

Abstract

"There are more than millions of motion picture which are present in today’s world. All of them provide entertainment for the user which help them spend their time. And yet there is no application out there to identify or recognize a motion picture which will be stumbled upon when browsing social media applications or when going through malls where the trailers are played aloud. The process of identifying motion pictures solely through audio waves and audio fingerprints has been a very challenging task in the multimedia processing. In this thesis, a novel approach has been used where the audio waves will be extracted to textual format and then the text would iterate through a SRT file using a recommendation model. In conclusion, this study provides a fresh method to the problem of motion picture identification using audio material. This method has the potential to be beneficial in a variety of applications, ranging from social media surfing to in-person movie watching, and it might open the way for additional study into the subject of motion picture identification."