Object Affordance Recognition on Object365 COCO-labels v1 (val)
57.53mAPSCL
Evaluation results
| Method | Links | |
|---|---|---|
| SCLSelf-training=true, Training Iterations=3M2022.03 | 57.53 | |
| SCLSelf-training=true, Training Iterations=0.8M2022.03 | 52.7 | |
| ATLOptimization Data=Standard2022.03 | 50.94 | |
| SCLSelf-training=true, Training Iterations=3M2022.03 | 46.32 | |
| SCL-Self-training=false, Training Iterations=0.8M2022.03 | 43.52 | |
| SCLSelf-training=true, Training Iterations=0.8M2022.03 | 42.05 | |
| SCL-Self-training=false, Training Iterations=0.8M2022.03 | 41.37 | |
| ATL*Optimization Data=COCO data2022.03 | 40.83 | |
| VCLTraining Iterations=0.8M2022.03 | 35.73 | |
| ATLOptimization Data=Standard2022.03 | 34.38 | |
| ATL*Optimization Data=COCO data2022.03 | 30.21 | |
| VCLTraining Iterations=0.8M2022.03 | 27.58 | |
| FCLTraining Iterations=0.8M2022.03 | 25.21 |