Semantic Object Segmentation on DAVIS 2016
0.82Jaccard MeanMaskContrast (Sup. Sal.)
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| MaskContrast (Sup. Sal.)Pre-training=ImageNet, Evaluation Protocol=Frozen representations, Saliency Type=Supervised2021.02 | 0.82 | 0.809 | |
| MaskContrast (Unsup. Sal.)Pre-training=ImageNet, Evaluation Protocol=Frozen representations, Saliency Type=Unsupervised2021.02 | 0.78 | 0.778 | |
| MoCo v2Pre-training=ImageNet, Evaluation Protocol=Frozen representations2021.02 | 0.771 | 0.772 |