Action spotting in soccer videos

What is it about?

We propose an efficient transformer-based system for action spotting in soccer videos. We first use the multi-scale vision transformer to extract features from the videos. Then we adopt a sliding window strategy to further utilize temporal features and enhanced temporal understanding. Finally, the features are input to NetVLAD++ model to obtain the final results. Our model can learn a hierarchy of robust representations and perform well.

Featured Image

Photo by Prapoth Panchuea on Unsplash

Photo by Prapoth Panchuea on Unsplash

Why is it important?

Action Spotting in the broadcast soccer game is important to understand salient actions and video summary applications. Our model can learn a hierarchy of robust representations and perform well. Our method achieves excellent results and outperforms the baseline and previous published works.

Perspectives