Image Retrieval On Par106K
評価指標
mAP
評価結果
このベンチマークにおける各モデルのパフォーマンス結果
モデル名 | mAP | Paper Title | Repository |
|---|---|---|---|
| DIR+QE* | 90.5% | Deep Image Retrieval: Learning global representations for image search | |
| Offline Diffusion | 96.2% | Efficient Image Retrieval via Decoupling Diffusion into Online and Offline Processing | |
| R-MAC+R+QE | 79.8% | Particular object retrieval with integral max-pooling of CNN activations | |
| DELF+FT+ATT+DIR+QE | 92.8% | Large-Scale Image Retrieval with Attentive Deep Local Features | |
| DELF+FT+ATT | 81.7% | Large-Scale Image Retrieval with Attentive Deep Local Features | |
| R-MAC | 75.7% | Particular object retrieval with integral max-pooling of CNN activations | |
| siaMAC+QE* | 78.3% | CNN Image Retrieval Learns from BoW: Unsupervised Fine-Tuning with Hard Examples |
0 of 7 row(s) selected.