Object Classification on CUB 50 classes
54.6Top-1 AccuracyPROTO. NETS
Evaluation Results
| Method | Links | |
|---|---|---|
| PROTO. NETSImage Features=GoogLeNet, Protocol=0-shot2017.03 | 54.6 | |
| DA-SJEImage Features=GoogLeNet, Protocol=0-shot2017.03 | 50.9 | |
| DS-SJEImage Features=GoogLeNet, Protocol=0-shot2017.03 | 50.4 | |
| SJEImage Features=GoogLeNet, Protocol=0-shot2017.03 | 50.1 | |
| SAMPLE CLUSTERINGImage Features=AlexNet, Protocol=0-shot2017.03 | 44.3 | |
| SJEImage Features=AlexNet, Protocol=0-shot2017.03 | 40.3 | |
| ALEFV dimensions=64K2015.03 | 26.9 | |
| ALEImage Features=Fisher, Protocol=0-shot2017.03 | 26.9 | |
| DAPFV dimensions=64K2015.03 | 12.3 |