Table 2. Quantitative results on the public dataset MICCAI EndoVis2017 (numbers in bold are the best indicator results).

Method	mIoU (%)	Mean F1 (%)	Mean Precision (%)	Mean Recall (%)
UNet	95.11	97.45	97.34	97.57
TransUNet	94.83	97.30	97.34	97.26
CGNet	93.42	96.52	96.51	96.53
SegFormer	93.85	96.76	97.70	96.82
GCNet	94.25	96.98	96.19	97.81
ContextNet	94.52	97.13	97.18	97.08
DABNet	95.37	97.59	97.51	97.68
FPENet	94.41	97.07	97.05	97.09
Ours	95.83	97.84	97.69	98.00