Cross-modal sequential scene retrieval on 3RScan 50 candidates (test)
81Rt@1SG2Loc
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| SG2Loc# of frames=50, Inference Time (ms)=75.0, Storage (MB)=5.42026.06 | 81 | 89 | 92 | |
| AnyLocInference Time (ms)=1451.1, Storage (MB)=5720.32026.06 | 81 | 87 | 90 | |
| SG2Loc# of frames=25, Inference Time (ms)=37.5, Storage (MB)=5.42026.06 | 78 | 87 | 91 | |
| SG2Loc# of frames=10, Inference Time (ms)=15.0, Storage (MB)=5.42026.06 | 75 | 85 | 89 | |
| SG2Loc# of frames=5, Inference Time (ms)=7.5, Storage (MB)=5.42026.06 | 72 | 84 | 88 | |
| SceneGraphLoc# of frames=1, Inference Time (ms)=1.5, Storage (MB)=5.42026.06 | 69 | 79 | 84 | |
| CVNetInference Time (ms)=311.1, Storage (MB)=239.12026.06 | 67 | 77 | 82 |