Instance Segmentation Propagation on DAVIS 2017
57.7J MeanOurs-track
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| Ours-trackSupervised=false, Pre-training Dataset=Kinetics2019.09 | 57.7 | 67.1 | 60 | 65.7 | |
| Joint-task Self-supervised LearningSupervised=false, Pre-training Dataset=Kinetics2019.09 | 56.8 | 65.7 | 59.5 | 65.1 | |
| OSVOSSupervised=true, Pre-training Dataset=ImageNet, DAVIS2019.09 | 56.6 | 63.8 | 63.9 | 73.8 | |
| SiamMaskSupervised=true, Pre-training Dataset=YouTube-VOS2019.09 | 54.3 | 62.8 | 58.5 | 67.5 | |
| ResNet-18Supervised=true, Pre-training Dataset=ImageNet, Configuration=3 blocks2019.09 | 49.4 | 52.9 | 55.1 | 56.6 | |
| Lai et al.Supervised=false, Pre-training Dataset=Kinetics2019.09 | 47.7 | — | 51.3 | — | |
| Wang et al.Supervised=false, Pre-training Dataset=VLOG, Resolution=480p2019.09 | 46.4 | 50.1 | 50 | 48 | |
| Wang et al.Supervised=false, Pre-training Dataset=VLOG, Resolution=400 x 4002019.09 | 43 | 43.7 | 42.6 | 41.3 | |
| mgPFFSupervised=false2019.09 | 42.2 | 41.8 | 46.9 | 44.4 | |
| ResNet-18Supervised=true, Pre-training Dataset=ImageNet, Configuration=4 blocks2019.09 | 40.2 | 36.1 | 42.5 | 36.6 | |
| DeepClusterSupervised=false, Pre-training Dataset=YFCC100M2019.09 | 37.5 | — | 33.2 | — | |
| Vondrick et al.Supervised=false, Pre-training Dataset=Kinetics2019.09 | 34.6 | 34.1 | 32.7 | 26.8 | |
| SIFT FlowSupervised=false2019.09 | 33 | — | 35 | — | |
| Transitive InvSupervised=false2019.09 | 32 | — | 26.8 | — |