Referring Expression Object Segmentation on RefCOCO UMD (testB)
70.62Accuracy (IoU > 0.5)full model
Evaluation Results
| Method | Links | |
|---|---|---|
| full modelInfo=Location, Joint training2019.10 | 70.62 | |
| MAttNetInfo=Attribute prediction, Attention module, Location, Relationships2019.10 | 69.99 | |
| + spatial-aware filtersInfo=Location2019.10 | 69.75 | |
| + caption-aware consistencyInfo=Joint training2019.10 | 69.36 | |
| + spatial coordsInfo=Location2019.10 | 68.54 | |
| VCInfo=Context2019.10 | 67.44 | |
| baseline2019.10 | 65.75 | |
| Yu et al.Info=Joint training2019.10 | 63.83 | |
| Liu et al.Info=Attribute prediction, Joint training2019.10 | 57.29 | |
| Nagaraja et al.Info=Context2019.10 | 56.4 | |
| Luo et al.Info=Joint training2019.10 | 55.18 |