Table 4 Ablation experiment
From: STEF: a Swin Transformer-Based Enhanced Feature Pyramid Fusion Model for Dongba character detection
FUSION | DB | Deforconv | Precision | Recall | F-measure |
|---|---|---|---|---|---|
\(\times\) | \(\times\) | \(\times\) | 76.62 | 72.33 | 74.42 |
\(\checkmark\) | \(\times\) | \(\times\) | 76.74 | 73.61 | 75.14 |
\(\times\) | \(\checkmark\) | \(\times\) | 84.29 | 81.23 | 82.73 |
\(\times\) | \(\checkmark\) | \(\checkmark\) | 80.62 | 83.97 | 82.26 |
\(\checkmark\) | \(\checkmark\) | \(\times\) | 85.11 | 84.13 | 84.62 |
\(\checkmark\) | \(\checkmark\) | \(\checkmark\) | 88.88 | 88.65 | 88.76 |
