Urban Perception Regression on Place Pulse Cross-Regional 2.0 (test)
9.3Cleanliness R^2UrbanFusion
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| UrbanFusionRegression Model=Linear, Input Modalities=SV, SV+RS, SV+OSM2025.10 | 9.3 | 15.2 | 24.8 | 30.1 | 24.7 | 23.4 | 21.2 | |
| GAIRRegression Model=Linear, Input Modalities=SV, SV+RS2025.10 | 9.1 | 14.5 | 24.7 | 28.5 | 23.4 | 22.5 | 20.4 | |
| GeoCLIPRegression Model=Linear, Input Modalities=SV2025.10 | 8.5 | 14.1 | 23.5 | 28.5 | 22.9 | 22.2 | 20 | |
| UrbanFusionRegression Model=MLP, Input Modalities=SV, SV+RS, SV+OSM2025.10 | 4.4 | 10.1 | 20.7 | 27 | 20.2 | 19.5 | 17 | |
| SatCLIP L40Regression Model=Linear, Input Modalities=RS2025.10 | 4.3 | 9.9 | 15 | 18.3 | 14.4 | 17.4 | 13.2 | |
| SatCLIP L10Regression Model=Linear, Input Modalities=RS2025.10 | 3.9 | 9.3 | 14.3 | 18.3 | 14.1 | 17.3 | 12.9 | |
| Identity (y ~ g(c))Regression Model=MLP2025.10 | 1.2 | 5.7 | 9.7 | 12.3 | 7.4 | 11.7 | 8 | |
| Identity (y ~ g(c))Regression Model=Linear2025.10 | 1 | 5.2 | 8.1 | 9.5 | 5.5 | 10 | 6.6 | |
| GeoCLIPRegression Model=MLP, Input Modalities=SV2025.10 | 0 | 3.9 | 15.6 | 21.3 | 12.6 | 15.7 | 11.5 | |
| SatCLIP L40Regression Model=MLP, Input Modalities=RS2025.10 | 0 | 1.7 | 10.7 | 12.4 | 8.4 | 11.5 | 7.4 | |
| GAIRRegression Model=MLP, Input Modalities=SV, SV+RS2025.10 | -1.3 | 5.2 | 16.6 | 20.1 | 14.9 | 15.9 | 11.9 | |
| SatCLIP L10Regression Model=MLP, Input Modalities=RS2025.10 | -1.8 | 2 | 9.4 | 12 | 8.3 | 11.9 | 7 |