Multi-task learning for jersey number recognition in Ice Hockey
This addresses player identification in sports videos, an incremental improvement for computer vision applications in sports analytics.
The paper tackles jersey number recognition in ice hockey videos by proposing a multi-task learning network that simultaneously learns holistic and digit-wise representations, achieving better performance than single-task networks.
Identifying players in sports videos by recognizing their jersey numbers is a challenging task in computer vision. We have designed and implemented a multi-task learning network for jersey number recognition. In order to train a network to recognize jersey numbers, two output label representations are used (1) Holistic - considers the entire jersey number as one class, and (2) Digit-wise - considers the two digits in a jersey number as two separate classes. The proposed network learns both holistic and digit-wise representations through a multi-task loss function. We determine the optimal weights to be assigned to holistic and digit-wise losses through an ablation study. Experimental results demonstrate that the proposed multi-task learning network performs better than the constituent holistic and digit-wise single-task learning networks.