Multimodal Spatial Reasoning on V-Star
61.28AccuracySpecFlow
Evaluation results
| Method | Links | |
|---|---|---|
| SpecFlowFLOPs(G)=13985.48, Lat.(s)=0.45, Mem.(GB)=41.222026.06 | 61.28 | |
| VoCoTFLOPs(G)=22334.01, Lat.(s)=0.71, Mem.(GB)=59.282026.06 | 59.87 | |
| SparseVLMFLOPs(G)=17802.00, Lat.(s)=0.54, Mem.(GB)=51.662026.06 | 57.36 | |
| CODIFLOPs(G)=14279.39, Lat.(s)=0.48, Mem.(GB)=42.652026.06 | 48.43 | |
| LightFastVFLOPs(G)=18802.51, Lat.(s)=0.60, Mem.(GB)=53.322026.06 | 46.83 | |
| PCCoTFLOPs(G)=13963.49, Lat.(s)=0.50, Mem.(GB)=42.312026.06 | 44.4 | |
| HeimaFLOPs(G)=14314.26, Lat.(s)=0.42, Mem.(GB)=40.622026.06 | 41.7 | |
| SoTFLOPs(G)=20319.23, Lat.(s)=0.66, Mem.(GB)=56.532026.06 | 38.06 |