Video Object Segmentation on DAVIS 21 (test)
70IoUPLM
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| PLMmode=fine-tuned, feature_layers=multiple, model_update=none (initial frame matching)2017.08 | 70 | 62 | 0.12 | |
| PLMpmode=without fine-tuning, feature_layers=multiple, model_update=none (initial frame matching)2017.08 | 68 | 58 | 0.12 | |
| OFLdescription=Baseline algorithm2017.08 | 67 | 63 | 0.25 | |
| PLMumode=fine-tuned, feature_layers=multiple, model_update=successive frame matching2017.08 | 63 | 52 | 0.16 | |
| BVSdescription=Baseline algorithm2017.08 | 59 | 58 | 0.13 | |
| JMPdescription=Baseline algorithm2017.08 | 56 | 53 | 0.18 | |
| NLCdescription=Baseline algorithm2017.08 | 55 | 52 | 0.26 | |
| HVSdescription=Baseline algorithm2017.08 | 54 | 52 | 0.24 | |
| SEAdescription=Baseline algorithm2017.08 | 50 | 47 | 0.25 | |
| PLMsmode=fine-tuned, feature_layers=single (last conv), model_update=none (initial frame matching)2017.08 | 23 | 16 | 0.13 |