Dense Captioning on Visual Genome (test)
36.29mAPCAG-Net
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| CAG-NetProposal Source=GT2019.04 | 36.29 | 0.316 | |
| T-LSTMProposal Source=GT2019.04 | 33.58 | 0.307 | |
| FCLNProposal Source=GT2019.04 | 27.03 | 0.305 | |
| FlexCap + GRIT BoxesFine-tuning=true, Boxes=GRIT Boxes2024.03 | 16.2 | — | |
| GRITFine-tuning=true2024.03 | 15.52 | — | |
| TDC+ROCSUFine-tuning=true2024.03 | 11.49 | — | |
| CAG-NetProposal Source=RPN2019.04 | 10.51 | 0.279 | |
| CAG-NetFine-tuning=true2024.03 | 10.51 | — | |
| COCGFine-tuning=true2024.03 | 9.82 | — | |
| T-LSTMProposal Source=RPN2019.04 | 9.31 | 0.275 | |
| JIVCFine-tuning=true2024.03 | 9.31 | — | |
| FCLNProposal Source=RPN2019.04 | 5.39 | 0.273 | |
| FCLNFine-tuning=true2024.03 | 5.39 | — |