Node Classification on COCO-SP LRGB (test)
43.98F1 ScoreNeuralWalker
Evaluation Results
| Method | Links | |
|---|---|---|
| NeuralWalker2024.06 | 43.98 | |
| GPSReference=[47]2024.06 | 38.84 | |
| GPS + TransformerParameter budget=500k2026.04 | 38.84 | |
| GPS + k-MIPParameter budget=500k2026.04 | 35.56 | |
| GPS + BigBirdParameter budget=500k2026.04 | 35.25 | |
| ExphormerBackbone=Exphormer2023.03 | 34.55 | |
| Exphormer2024.06 | 34.55 | |
| ExphormerParameter budget=500k2026.04 | 34.55 | |
| GPS2022.05 | 34.12 | |
| GraphGPSBackbone=GraphGPS2023.03 | 34.12 | |
| GPSReference=[42]2024.06 | 34.12 | |
| GatedGCN2024.06 | 29.22 | |
| GatedGCNParameter budget=500k2026.04 | 29.22 | |
| GPS + PerformerParameter budget=500k2026.04 | 27.7 | |
| GatedGCN2022.05 | 26.41 | |
| GatedGCNBackbone=GatedGCN2023.03 | 26.41 | |
| Transformerpositional_encoding=LapPE2022.05 | 26.18 | |
| Transformer+LapPEBackbone=Transformer, Positional Encoding=LapPE2023.03 | 26.18 | |
| SANpositional_encoding=LapPE2022.05 | 25.92 | |
| SAN+LapPEBackbone=SAN, Positional Encoding=LapPE2023.03 | 25.92 | |
| GatedGCNstructural_encoding=RWSE2022.05 | 25.74 | |
| GatedGCN+RWSEBackbone=GatedGCN, Positional Encoding=RWSE2023.03 | 25.74 | |
| SANstructural_encoding=RWSE2022.05 | 24.34 | |
| SAN+RWSEBackbone=SAN, Positional Encoding=RWSE2023.03 | 24.34 | |
| GIN2024.06 | 21.25 | |
| GINEParameter budget=500k2026.04 | 21.25 | |
| GATParameter budget=500k2026.04 | 18.86 | |
| GINE2022.05 | 13.39 | |
| GINEBackbone=GINE2023.03 | 13.39 | |
| GCN2024.06 | 13.38 | |
| GCNParameter budget=500k2026.04 | 13.38 | |
| GCN2022.05 | 8.41 | |
| GCNBackbone=GCN2023.03 | 8.41 |