Visual Repetition Counting on RepCount benchmark split
0.013MAERepNet
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| RepNetEncoder=R2D502024.03 | 0.013 | 0.995 | — | — | |
| ESCountsEncoder=ViT-B2024.03 | 0.213 | 4.455 | 0.245 | 0.563 | |
| SkimFocusEncoder=SwinB2024.03 | 0.249 | — | — | 0.517 | |
| ESCountsEncoder=SwinT2024.03 | 0.298 | 6.905 | 0.183 | 0.403 | |
| MFLEncoder=SwinT, multi-modal=true2024.03 | 0.384 | — | — | 0.543 | |
| TransRACEncoder=SwinT, reproduced=true2024.03 | 0.443 | 9.13 | 0.085 | 0.291 | |
| DeTRCEncoder=ViT-B2024.03 | 0.625 | — | — | 0.204 |