arXiv Analytics

Sign in

arXiv:1705.01781 [cs.CV]AbstractReferencesReviewsResources

Am I Done? Predicting Action Progress in Videos

Federico Becattini, Tiberio Uricchio, Lamberto Ballan, Lorenzo Seidenari, Alberto Del Bimbo

Published 2017-05-04Version 1

In this paper we introduce the problem of predicting action progress in untrimmed videos. We argue that this is an extremely important task because, on the one hand, it can be valuable for a wide range of applications and, on the other hand, it facilitates better action detection results. To solve this problem we introduce a novel approach, named ProgressNet, capable of predicting when an action takes place in a video, where it is located within the frames, and how far it has progressed during its execution. Motivated by the recent success obtained from the interaction of Convolutional and Recurrent Neural Networks, our model is based on a combination of the well known Faster R-CNN framework, to make framewise predictions, and LSTM networks, to estimate action progress through time. After introducing two evaluation protocols for the task at hand, we demonstrate the capability of our model to effectively predict action progress on a subset of 11 classes from UCF-101, all of which exhibit strong temporal structure. Moreover, we show that this leads to state-of-the-art spatio-temporal localization results.