The rise of Artificial Intelligence (AI) is changing the creative industry and offering new solutions to simplify traditional and complex tasks. This program uses AI to automate video editing, focusing on removing silence and bad takes. The workflow starts by extracting audio from a video file and then producing a dialogue text. This text is processed using the model, which edits the text by removing fumbles, repetitions, and other errors, resulting in a clearer and better version. In parallel, word-level alignment is performed on the text to accurately map each word to its timestamp in the video. The clean text generated by the model is used to filter and remove unwanted parts from the scanned transcript. By doing this, you will always have relevant and quality content. This method not only increases the accuracy and speed of video editing but also reduces the manual effort required, allowing content creators to focus on the creative aspects of their work. Incorporating artificial intelligence into this approach shows the potential of automation to transform video editing, making it more efficient and accessible.
Department of Electronics and Computer Engineering
Pulchowk Campus, Institute of Engineering, Tribhuvan University