Image Captioning on Flickr 30k (BLEU-1 to 4, METEOR, ROUGE, CIDEr, SPICE)
0.364BLEU-4Proposed Model
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Proposed ModelComplexity=near-linear O(nk)2026.06 | 0.364 | 0.733 | 0.438 | 0.267 | 0.368 | 0.568 | 0.723 | 0.199 | |
| Scene Graph2026.06 | 0.338 | — | — | — | 0.243 | 0.479 | 0.682 | 0.184 | |
| Multimodal2026.06 | 0.333 | 0.758 | 0.587 | 0.445 | 0.245 | — | — | — | |
| ADA2026.06 | 0.324 | 0.794 | 0.596 | 0.437 | 0.283 | 0.608 | 0.946 | 0.188 | |
| GNN-Visual2026.06 | 0.322 | 0.752 | 0.541 | 0.406 | 0.245 | 0.525 | 0.646 | — | |
| GAT2026.06 | 0.308 | 0.744 | 0.567 | 0.418 | 0.234 | — | 0.68 | — | |
| TopDown2026.06 | 0.308 | 0.737 | — | — | 0.232 | — | 0.706 | 0.176 | |
| RFT2026.06 | 0.306 | 0.716 | — | — | 0.237 | 0.516 | 0.733 | 0.174 | |
| High-Order2026.06 | 0.301 | 0.732 | — | — | 0.229 | 0.508 | 0.65 | 0.168 | |
| SCST2026.06 | 0.301 | 0.734 | — | — | 0.226 | — | 0.693 | 0.168 | |
| AoA2026.06 | 0.297 | 0.728 | — | — | 0.223 | 0.501 | 0.643 | 0.166 | |
| TBDATN2026.06 | 0.295 | 0.765 | 0.589 | 0.445 | 0.258 | 0.602 | 0.702 | — | |
| DRET2026.06 | 0.288 | 0.705 | 0.527 | 0.39 | 0.226 | 0.498 | 0.636 | — | |
| TAKG2026.06 | 0.265 | 0.683 | — | — | 0.219 | 0.48 | 0.575 | 0.163 |