Detecting events and key actors in multi-person videos
This addresses the problem of event recognition in crowded scenes for video analysis applications, representing an incremental improvement with a novel attention mechanism.
The paper tackles multi-person event recognition in videos by proposing a model that detects events and identifies key actors without explicit annotations, outperforming state-of-the-art methods on a new basketball dataset with 257 games and 14K annotations across 11 event classes.
Multi-person event recognition is a challenging task, often with many people active in the scene but only a small subset contributing to an actual event. In this paper, we propose a model which learns to detect events in such videos while automatically "attending" to the people responsible for the event. Our model does not use explicit annotations regarding who or where those people are during training and testing. In particular, we track people in videos and use a recurrent neural network (RNN) to represent the track features. We learn time-varying attention weights to combine these features at each time-instant. The attended features are then processed using another RNN for event detection/classification. Since most video datasets with multiple people are restricted to a small number of videos, we also collected a new basketball dataset comprising 257 basketball games with 14K event annotations corresponding to 11 event classes. Our model outperforms state-of-the-art methods for both event classification and detection on this new dataset. Additionally, we show that the attention mechanism is able to consistently localize the relevant players.