Task-incremental Learning Accuracy on CelebA 20 binary attributes
88.2Average AccuracyMulti-task
Evaluation results
| Method | Links | |
|---|---|---|
| Multi-taskBackbone=AlexNet, Task Setting=Joint, Number of runs=32026.04 | 88.2 | |
| DRCLBackbone=AlexNet, Task Setting=Joint, Number of runs=32026.04 | 87.8 | |
| SGPBackbone=AlexNet, Task Setting=Joint, Number of runs=32026.04 | 87.3 | |
| IBPCLBackbone=AlexNet, Task Setting=Joint, Number of runs=32026.04 | 87.1 | |
| UPGDBackbone=AlexNet, Task Setting=Joint, Number of runs=32026.04 | 87 | |
| SB-MCLBackbone=AlexNet, Task Setting=Joint, Number of runs=32026.04 | 86.9 | |
| BANBackbone=AlexNet, Task Setting=Joint, Number of runs=32026.04 | 86.5 | |
| EWCBackbone=AlexNet, Task Setting=Joint, Number of runs=32026.04 | 86.3 | |
| UCLBackbone=AlexNet, Task Setting=Joint, Number of runs=32026.04 | 86.3 | |
| POCLBackbone=AlexNet, Task Setting=Joint, Number of runs=32026.04 | 85.7 | |
| BaselineBackbone=AlexNet, Task Setting=Joint, Number of runs=32026.04 | 84.6 | |
| WSNBackbone=AlexNet, Task Setting=Joint, Number of runs=32026.04 | 84.6 | |
| A-GEMBackbone=AlexNet, Task Setting=Joint, Number of runs=32026.04 | 83.7 | |
| HATBackbone=AlexNet, Task Setting=Joint, Number of runs=32026.04 | 82.5 | |
| TAGBackbone=AlexNet, Task Setting=Joint, Number of runs=32026.04 | 78.3 |