Profanity Detection and Removal in Videos using Machine Learning

With the increase in content being uploaded daily on video sharing platforms, the need to moderate the material uploaded becomes mandatory. Traditional profane removal mechanisms follow a prolonged approach. This is not recommended since the audience already had access to the media before it was identified of having potential profane content. There has been extensive work in the domain focusing on profanity removal from text. There arises a need for automated video moderation technique to handle the media which is being uploaded at the initial stage itself so as to provide a profane free platform to the viewers which ranges from kids to the adults. With a view to detect and remove profanity from the video-sharing platforms, this paper suggests an approach to silence the audio and pixelate the lips present if any in the video-segments containing profanity using machine learning.