Attribute Prediction on Clothing dataset 8 (test)
93.12Accuracy (Color)MG-CNN
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| MG-CNNBackbone=AlexNet2022.07 | 93.12 | 95.37 | 88.65 | 91.93 | 92.82 | |
| Label2LabelBackbone=AlexNet, Attribute query network layers=2-layer Transformer decoder2022.07 | 92.73 | 96.82 | 88.2 | 90.88 | 92.87 | |
| AQNBackbone=AlexNet2022.07 | 91.98 | 96.37 | 88.19 | 89.89 | 92.29 | |
| M-CNNBackbone=AlexNet2022.07 | 91.72 | 94.26 | 87.96 | 91.51 | 91.7 | |
| Meng et al.Backbone=AlexNet2022.07 | 91.64 | 96.81 | 89.25 | 89.53 | 92.39 | |
| FC HeadBackbone=AlexNet2022.07 | 91.39 | 96.07 | 87 | 88.21 | 91.57 | |
| S-CNNBackbone=AlexNet2022.07 | 90.5 | 92.9 | 87 | 89.57 | 90.43 | |
| MG-CNNFramework=Multi-task Learning (MTL) with group encodings2016.01 | 0.9312 | 0.9537 | 0.8865 | 0.9193 | 0.9282 | |
| M-CNNFramework=Multi-task Learning (MTL) without group encoding2016.01 | 0.9172 | 0.9426 | 0.8796 | 0.9151 | 0.917 | |
| S-CNNModel=Single-task CNN, Training=Fine-tuned individual models2016.01 | 0.905 | 0.929 | 0.87 | 0.8957 | 0.9043 | |
| CRF2016.01 | 0.85 | 0.8433 | 0.8125 | 0.825 | 0.8395 | |
| M-extractFeature Extraction=Pre-trained CNN extracted features, Framework=Multi-task Learning (MTL)2016.01 | 0.8498 | 0.8989 | 0.8141 | 0.8103 | 0.8529 | |
| S-extractFeature Extraction=Pre-trained CNN [25], Classifier=Single SVM tasks2016.01 | 0.8184 | 0.8207 | 0.6751 | 0.6925 | 0.7831 | |
| CFBaseline=Combined features model with no pose2016.01 | 0.81 | 0.8208 | 0.7763 | 0.785 | 0.8048 |