Table 6 Comparison of acc metrics between the proposed method and state-of-the-art methods in robustness experiments
From: A Mamba based vision transformer for fine grained image segmentation of mural figures
Acc | ||||||
|---|---|---|---|---|---|---|
Methods | Other | Face | Headwear | Body | mAcc | aAcc |
PSPNet | 92.53 | 83.60 | 47.98 | 88.45 | 78.14 | 89.72 |
OCRNet | 89.04 | 92.94 | 54.89 | 88.51 | 81.35 | 87.87 |
Segmenter | 92.48 | 70.92 | 48.22 | 86.92 | 74.64 | 88.98 |
SegFormer | 92.8 | 89.97 | 53.05 | 82.43 | 79.56 | 88.49 |
DDRNet | 88.44 | 73.53 | 39.29 | 83.54 | 71.20 | 85.12 |
SAN | 81.71 | 88.36 | 55.75 | 75.71 | 75.38 | 79.31 |
Ours | 93.89 | 87.72 | 56.42 | 79.84 | 79.47 | 88.53 |
