Semantic Segmentation on Cityscapes 80-20 (val) benchmark leaderboard · SOTA2 Research

Semantic Segmentation on Cityscapes 80-20 (val)

84.05AccuracyLOOPE

Evaluation Results

Method	Links
LOOPE2025.04		84.05	82.26	87.31	32.21	86.37
LOOPE2025.04		83.63	81.37	86.17	7.44	89.56
LFF(Fourier)2025.04		83.2	81.41	86.46	26.6	85.51
LFF(Fourier)2025.04		82.69	80.41	85.2	7.83	85.64
1D Sinusoid2025.04		81.82	80.59	85.79	18.37	84.74
Learnable2025.04		81.48	80.46	85.73	15.33	84.68
Learnable2025.04		81.35	74.41	79.26	3.88	88.87
1D Sinusoid2025.04		80.96	78.33	82.51	3.55	91.99
2D Sinusoid2025.04		80.93	79.9	85.13	9.58	84.08
No PE2025.04		80.84	79.56	84.53	3.48	88.75
2D Sinusoid2025.04		80.74	78.26	82.46	3.04	91.88
CPE2025.04		80.67	78.42	82.71	3.31	87.26
No PE2025.04		79.55	79.13	84.48	6.23	83.34
CPE2025.04		79.19	78.71	84	8.8	82.89