Localization-aware captioning on Simulated Radar Dataset (10-20m)
93.8PrecisionSG-CLIP
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| SG-CLIPalpha=1.02025.11 | 93.8 | 93 | 93.4 | |
| SG-CLIPalpha=16.02025.11 | 92.9 | 91.6 | 92.2 | |
| SG-CLIPalpha=4.02025.11 | 92.7 | 91.8 | 92.2 | |
| Vanilla CLIP2025.11 | 90 | 81.7 | 84.4 |