Sports Ball Detection And Tracking On Tennis
Metrics
Accuracy (%)
Average Precision (%)
F1 (%)
Results
Performance results of various models on this benchmark
Model Name | Accuracy (%) | Average Precision (%) | F1 (%) | Paper Title | Repository |
|---|---|---|---|---|---|
| DeepBall | 32.3 | 47.0 | 47.4 | DeepBall: Deep Neural-Network Ball Detector | - |
| WASB (Step=1) | 91.8 | 94.2 | 95.6 | Widely Applicable Strong Baseline for Sports Ball Detection and Tracking | - |
| MonoTrack | 85.9 | 87.3 | 92.1 | MonoTrack: Shuttle trajectory reconstruction from monocular badminton video | - |
| TrackNetV2 | 81.4 | 80.6 | 89.4 | TrackNetV2: Efficient Shuttlecock Tracking Network | |
| WASB (Step=3) | 89.0 | 91.0 | 94.0 | Widely Applicable Strong Baseline for Sports Ball Detection and Tracking | - |
| ResTrackNetV2 | 82.8 | 81.7 | 90.3 | Widely Applicable Strong Baseline for Sports Ball Detection and Tracking | - |
| BallSeg | 57.5 | 56.8 | 71.7 | Real-time CNN-based Segmentation Architecture for Ball Detection in a Single View Setup | - |
| DeepBall-Large | 31.6 | 35.1 | 46.7 | Widely Applicable Strong Baseline for Sports Ball Detection and Tracking | - |
0 of 8 row(s) selected.