Text-based Video Segmentation on A2D Sentences
41.2mAP (0.5:0.95)AAMN
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| AAMNInput=RGB+Flow, Speed=546.5 ms2020.11 | 41.2 | 68.9 | 64.2 | 54.5 | 32.4 | 340 | 63.4 | 55.6 | |
| AAMNInput=RGB, Speed=446.6 ms2020.11 | 34.8 | 65.2 | 58 | 44.5 | 20.8 | 160 | 58.8 | 51.3 | |
| CMDyInput=RGB+Flow, Speed=201.3 ms2020.11 | 34.2 | 63.3 | 55 | 42.8 | 25.7 | 490 | 62.9 | 53.8 | |
| CMDyInput=RGB, Speed=125.3 ms2020.11 | 33.3 | 60.7 | 52.5 | 40.5 | 23.5 | 450 | 62.3 | 53.1 | |
| VT-CapsuleInput=RGB2020.11 | 30.3 | 52.6 | 45 | 34.5 | 20.7 | 360 | 56.8 | 46 | |
| ACGAInput=RGB+Flow, Speed=184.8 ms2020.11 | 28.7 | — | — | — | — | — | 60.6 | 50.3 | |
| ACGAInput=RGB, Speed=108.9 ms2020.11 | 27.4 | 55.7 | 45.9 | 31.9 | 16 | 200 | 60.1 | 49 | |
| Gavrilyuk et al.Input=RGB+Flow, Speed=174.7 ms2020.11 | 21.5 | 50 | 37.6 | 23.1 | 9.4 | 40 | 55.1 | 42.6 | |
| Gavrilyuk et al.Input=RGB, Speed=95.6 ms2020.11 | 19.8 | 47.5 | 34.7 | 21.1 | 8 | 20 | 53.6 | 42.1 | |
| Li et al.Input=RGB2020.11 | 16.3 | 38.7 | 29 | 17.5 | 6.6 | 10 | 51.5 | 35.4 | |
| Hu et al.Input=RGB, Speed=60.6 ms2020.11 | 13.2 | 34.8 | 23.6 | 13.3 | 3.3 | 10 | 47.4 | 35 |