Table 1 Comparison of IoU metrics between the proposed method and state-of-the-art approaches for mural figures on the test set
From: A Mamba based vision transformer for fine grained image segmentation of mural figures
IoU | ||||||
|---|---|---|---|---|---|---|
Method | Other | Face | Headwear | Bracelet | Armlet | mIoU |
PSPNet | 99.32 | 90.76 | 68.71 | 28.24 | 34.24 | 64.25 |
OCRNet | 99.24 | 87.39 | 65.49 | 29.76 | 35.95 | 63.57 |
Segmenter | 98.90 | 83.73 | 50.17 | 17.22 | 33.71 | 56.75 |
SegFormer | 99.34 | 89.47 | 69.52 | 35.49 | 37.14 | 66.19 |
DDRNet | 99.15 | 88.25 | 59.01 | 26.86 | 28.63 | 60.38 |
SAN | 99.14 | 88.75 | 59.53 | 28.17 | 24.24 | 59.96 |
Ours | 99.35 | 89.97 | 67.71 | 36.15 | 48.95 | 68.42 |
