Action Unit Detection on BP4D (3-fold cross val)
63Average F1AUNets
Evaluation results
| Method | Links | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| AUNetsarchitecture=RGB+OF Horizontal concatenation, temporal_window=7 (median), evaluation_protocol=3-fold cross validation2017.04 | 63 | 53.4 | 44.7 | 55.8 | 79.2 | 78.2 | 83.1 | 88.4 | 66.6 | 47.5 | 62 | 47.3 | 49.7 | |
| AU R-CNNBackbone=ResNet2018.12 | 63 | — | — | — | — | — | — | — | — | — | — | — | — | |
| AU R-CNNBackbone=VGG-162018.12 | 60.7 | — | — | — | — | — | — | — | — | — | — | — | — | |
| AU R-CNNBackbone=VGG-192018.12 | 60.3 | — | — | — | — | — | — | — | — | — | — | — | — | |
| ROI-Nets2018.12 | 56.4 | — | — | — | — | — | — | — | — | — | — | — | — | |
| EACevaluation_protocol=3-fold cross validation2017.04 | 55.9 | 39 | 35.2 | 48.6 | 76.1 | 72.9 | 81.9 | 86.2 | 58.8 | 37.5 | 59.1 | 35.9 | 35.8 | |
| EAC-Net2018.12 | 55.9 | — | — | — | — | — | — | — | — | — | — | — | — | |
| OFS-CNN2018.12 | 53.7 | — | — | — | — | — | — | — | — | — | — | — | — | |
| MSTCevaluation_protocol=3-fold cross validation2017.04 | 53.2 | 31.4 | 31.1 | 71.4 | 63.3 | 77.1 | 45 | 82.6 | 72.9 | 34 | 53.9 | 38.5 | 37 | |
| CNN+LSTM2018.12 | 53.2 | — | — | — | — | — | — | — | — | — | — | — | — | |
| FERA2018.12 | 51.7 | — | — | — | — | — | — | — | — | — | — | — | — | |
| CPM2018.12 | 50 | — | — | — | — | — | — | — | — | — | — | — | — | |
| DRMLevaluation_protocol=3-fold cross validation2017.04 | 48.3 | 36.4 | 41.8 | 43 | 55 | 67 | 66.3 | 65.8 | 54.1 | 33.2 | 48 | 31.7 | 30 | |
| DRML2018.12 | 48.3 | — | — | — | — | — | — | — | — | — | — | — | — | |
| JPMLevaluation_protocol=3-fold cross validation2017.04 | 45.9 | 32.6 | 25.6 | 37.4 | 42.3 | 50.5 | 72.2 | 74.1 | 65.7 | 38.1 | 40 | 30.4 | 42.3 | |
| JPML2018.12 | 45.9 | — | — | — | — | — | — | — | — | — | — | — | — | |
| LSVM2018.12 | 35.3 | — | — | — | — | — | — | — | — | — | — | — | — |