Some “fail” videos leave us asking ourselves if it was intentional or not. This new annotated video dataset is abundant with unintentional action, which has been collected by crawling publicly available “fail” videos from the web. The dataset is both large (over 50 hours of video) and diverse (covering hundreds of scenes and activities). The videos are annotated videos with the temporal location at which the video transitions from intentional to unintentional action. Now we will be able to know, I guess.

Masashi Kawamura: “A fun research and experiment into predicting failure using deep learning. These types of fundamental research are very inspiring to check out.”