Context According to online sources, basic video editing can take 30 minutes to an hour for each minute. This discourages users to produce periodic content. Currently, Automatic Video Editing is limited to video enhances and simple mechanisms such as silence or clapping detection have been applied in commercial tools. In order to have smarter detectors, we need to recognize complex actions such as bloopers. Bloopers are mistakes made by a person on screen. Monologues are a perfect setting for blooper detection because of fixed conditions. Monologues involve one person on screen and a fixed camera position. In this video, we collected videos from youtube and separate short video fragments into blooper and no blooper video clips. This can help programmers to train models able to detect bloopers in full-length videos. Content Number of videos: 600 videos Time length: Between 1 and 3 seconds per video clip Batches: Train, Validation, and Test batches Train: 464 videos Test: 66 videos Validation: 66 videos Categories: 2 categories 0 Blooper 1 No blooper The videos are multilingual. Both genders are represented. Acknowledgements West Virginia University professors. Inspiration Help us to make it grow.