Table 3: Comparison of bowel inflammation detection results between the baseline and the proposed methods for All GT.
All GT | mAP (%) | Precision (%) | Recall (%) |
Methods | @10 | @50 | @10 | @50 | @10 | @50 |
SSSL (Baseline) | 68.83 | 62.27 | 66.17 | 58.16 | 70.54 | 65.46 |
MSSL (s = 1) | 63.60 | 56.48 | 59.84 | 51.31 | 65.84 | 60.33 |
MSSL (s = 2) | 65.30 | 58.03 | 61.84 | 53.77 | 67.44 | 61.42 |
MSSL (s = 3) | 66.95 | 60.90 | 64.56 | 57.19 | 68.71 | 63.77 |
MSML (s = 1) | 69.27 | 63.06 | 67.10 | 59.68 | 70.90 | 65.85 |
MSML (s = 2) | 67.45 | 61.52 | 65.56 | 58.23 | 69.41 | 64.45 |
MSML (s = 3) | 67.65 | 62.11 | 65.93 | 59.27 | 69.18 | 64.71 |