Photo Aesthetics Classification on AVA (test)
79.08AccuracyMTRLCNN
Evaluation results
| Method | Links | |
|---|---|---|
| MTRLCNNdelta=0, Backbone=ResNet, Fine-tuning=true2016.04 | 79.08 | |
| MTRLCNNdelta=0, Backbone=VGG Net, Fine-tuning=true2016.04 | 78.46 | |
| MTRLCNNdelta=1, Backbone=ResNet, Fine-tuning=true2016.04 | 77.71 | |
| MTRLCNNdelta=1, Backbone=VGG Net, Fine-tuning=true2016.04 | 77.41 | |
| MNA-CNNdelta=0, Backbone=VGG Net, Fine-tuning=true2016.04 | 77.4 | |
| MTRLCNNdelta=0, Backbone=AlexNet, Fine-tuning=true2016.04 | 77.35 | |
| Reg+Rank+Att+ContLoss=Regression + Ranking, Features=Attributes + Content2016.06 | 77.33 | |
| MTRLCNNdelta=1, Backbone=AlexNet, Fine-tuning=true2016.04 | 76.8 | |
| MNA-CNNdelta=1, Backbone=VGG Net, Fine-tuning=true2016.04 | 76.5 | |
| MTCNN #1delta=02016.04 | 76.15 | |
| MTCNN #1delta=12016.04 | 75.9 | |
| Reg+Rank+AttLoss=Regression + Ranking, Features=Attributes2016.06 | 75.48 | |
| RDCNN_semantic2016.06 | 75.42 | |
| DMA_AlexNet_FTBackbone=AlexNet2016.06 | 75.41 | |
| DMA-Netdelta=02016.04 | 75.41 | |
| Reg+AttLoss=Regression, Features=Attributes2016.06 | 75.32 | |
| STCNNdelta=12016.04 | 75.13 | |
| RDCNN2016.06 | 74.46 | |
| DMA2016.06 | 74.46 | |
| RDCNNdelta=02016.04 | 74.46 | |
| RDCNNdelta=12016.04 | 73.7 | |
| Reg+Rank+ContLoss=Regression + Ranking, Features=Content2016.06 | 73.37 | |
| DCNN2016.06 | 73.25 | |
| SPP2016.06 | 72.85 | |
| STCNNdelta=02016.04 | 72.19 | |
| RegLoss=Regression2016.06 | 72.04 | |
| AlexNet_FT_ConfBackbone=AlexNet2016.06 | 71.52 | |
| Reg+RankLoss=Regression + Ranking2016.06 | 71.5 | |
| SCNNdelta=02016.04 | 71.2 | |
| SCNNdelta=12016.04 | 68.63 | |
| Murray et al.2016.06 | 68 | |
| Murraydelta=12016.04 | 67 | |
| Murraydelta=02016.04 | 66.7 |