Instance Segmentation on NYUD v2 (test)
39.3Mean APGSPN
Evaluation results
| Method | Links | ||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| GSPN2018.12 | 39.3 | 62.8 | 51.4 | 35.1 | 11.4 | 54.6 | 45.8 | 38 | 22.9 | 43.3 | 8.4 | 36.8 | 18.3 | 58.1 | 42 | 45.4 | 54.8 | 29.1 | 20.8 | 67.5 | |
| SGPN-CNN2018.12 | 33.6 | 45.3 | 62.5 | 43.9 | 0 | 45.6 | 40.7 | 30 | 20.2 | 42.6 | 8.8 | 28.2 | 15.5 | 43 | 30.4 | 51.4 | 58.9 | 25.6 | 6.6 | 39 | |
| R-CNN (HHA)Modality=RGB-D, Encoding=HHA (geocentric embedding)2014.07 | 32.1 | 18.9 | 66.1 | 10.2 | 1.5 | 35.5 | 32.8 | 10.2 | 22.8 | 33.7 | 38.3 | 35.5 | 53.3 | 42.7 | 31.5 | 34.4 | 40.7 | 14.3 | 37.4 | 50.5 | |
| MRCNN*depth information=true2018.12 | 31.5 | 24.7 | 66.3 | 20.1 | 1.4 | 44.9 | 43.9 | 6.8 | 16.6 | 29.5 | 22.1 | 29.2 | 29.3 | 36.9 | 34.6 | 37.1 | 48.4 | 26.6 | 21.9 | 58.5 | |
| MRCNNdepth information=false2018.12 | 29.3 | 26.3 | 54.1 | 23.4 | 3.1 | 39.3 | 34 | 6.2 | 17.8 | 23.7 | 23.1 | 31.1 | 35.1 | 25.4 | 26.6 | 36.4 | 47.1 | 21 | 23.3 | 58.8 | |
| regionsegmentation_primitive=region baseline2014.07 | 28.1 | 32.4 | 54.9 | 9.4 | 1.1 | 27 | 21.4 | 8.9 | 20.3 | 29 | 37.1 | 26.3 | 48.3 | 38.6 | 33.1 | 30.9 | 30.5 | 10.2 | 33.7 | 39.9 | |
| fg masksegmentation_primitive=foreground mask2014.07 | 28 | 14.7 | 59.9 | 8.9 | 1.3 | 29.2 | 5.4 | 7.2 | 22.6 | 33.2 | 38.1 | 31.2 | 54.8 | 39.4 | 32.1 | 32 | 36.2 | 11.2 | 37.4 | 37.5 | |
| boxsegmentation_primitive=box detections2014.07 | 14 | 5.9 | 40 | 4.1 | 0.7 | 5.5 | 0.5 | 3.2 | 14.5 | 26.9 | 32.9 | 1.2 | 40.2 | 11.1 | 6.1 | 9.4 | 13.6 | 2.6 | 35.1 | 11.9 |