Fine-grained Image Classification on Dogs
92.75AccuracyRelational Proxy
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Relational Proxy2022.10 | 92.75 | 1.45 | 1.29 | |
| TransFG2022.10 | 92.3 | 1 | 0.9 | |
| Relation-Agnostic Encoder2022.10 | 91.3 | — | — | |
| Ours (multi scale)Backbone=ResNet-50, Resolution=448, #Parameters=23.9M, scale_mode=multi2019.12 | 87.7 | — | — | |
| RACNNBackbone=VGG-19, Resolution=448, #Parameters=129M2019.12 | 87.3 | — | — | |
| Ours (single scale)Backbone=ResNet-50, Resolution=448, #Parameters=23.9M, scale_mode=single2019.12 | 87.1 | — | — | |
| MAMCBackbone=ResNet-50, Resolution=448, #Parameters=434M2019.12 | 85.2 | — | — | |
| PCBackbone=ResNet-50, #Parameters=23.9M2019.12 | 83.8 | — | — | |
| MaxEntBackbone=ResNet-50, #Parameters=23.9M2019.12 | 83.6 | — | — |