RMS-Net: Regression and Masking for Soccer Event Spotting
This work addresses event detection in soccer videos, providing incremental improvements for sports analytics applications.
The paper tackles the action spotting task in soccer videos by proposing RMS-Net, a lightweight modular network that predicts event labels and temporal offsets using shared features, enhanced with data balancing and masking strategies. It achieves a 3 Average-mAP point improvement over state-of-the-art on the SoccerNet dataset with standard features and over 10 points with fine-tuning.
The recently proposed action spotting task consists in finding the exact timestamp in which an event occurs. This task fits particularly well for soccer videos, where events correspond to salient actions strictly defined by soccer rules (a goal occurs when the ball crosses the goal line). In this paper, we devise a lightweight and modular network for action spotting, which can simultaneously predict the event label and its temporal offset using the same underlying features. We enrich our model with two training strategies: the first one for data balancing and uniform sampling, the second for masking ambiguous frames and keeping the most discriminative visual cues. When tested on the SoccerNet dataset and using standard features, our full proposal exceeds the current state of the art by 3 Average-mAP points. Additionally, it reaches a gain of more than 10 Average-mAP points on the test set when fine-tuned in combination with a strong 2D backbone.