Generation of Video Subtitles for All Video Creators

With closed captions, those with hearing impairments may easily follow the information in videos, making them an essential accessibility tool. Nevertheless, the process of making subtitles can be costly and time-consuming. Making video subtitles automatically is a promising way to lower the cost and increase accessibility of subtitles. In this project, we create and assess a system for automatically captioning videos. Our method converts audio from video using the most sophisticated speech recognition model. After that, we apply a range of post-processing methods to enhance the quality of the subtitles, including mistake correction, text segmentation, and timestamp addition. We test our system on a video dataset with numerous speakers, diverse accents, and background noise, among other auditory circumstances. We also do a comparison between our method and other cutting-edge video subtitling systems. Our findings demonstrate that our technology performs competitively across the board. We also discovered that our algorithm can produce understandable, crisp subtitles. We are confident that our system can have a major impact on the automatic video captioning industry.