Task-incremental Image Classification on Tiny-ImageNet 200 10 (test)
67.2Task 1 ScoreEFT-a8b16
Evaluation results
| Method | Links | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| EFT-a8b16Backbone=VGG-16, Parameter Increase (%)=3.1%2021.03 | 67.2 | 62.5 | 69.4 | 62.6 | 68.3 | 69.6 | 59 | 67.8 | 71.5 | 70.1 | 66.8 | |
| PackNetBackbone=VGG-162021.03 | 52.5 | 49.7 | 56.5 | 59.8 | 55 | 64.7 | 61.7 | 55.9 | 55.2 | 52.5 | 56.4 | |
| FreezingBackbone=VGG-162021.03 | 51.7 | 36.4 | 39.5 | 41.7 | 42.9 | 46.2 | 45.7 | 41.1 | 41.2 | 40.9 | 42.7 | |
| IMM-modeBackbone=VGG-162021.03 | 50.6 | 38.5 | 44.7 | 49.2 | 47.5 | 51.9 | 53.7 | 47.7 | 50 | 48.7 | 48.3 | |
| TFMBackbone=VGG-162021.03 | 48.2 | 47.7 | 56.7 | 58.2 | 54.8 | 62.2 | 61.5 | 57.3 | 58.5 | 54.8 | 56 | |
| HATBackbone=VGG-162021.03 | 46.8 | 49.1 | 55.8 | 58 | 53.7 | 61 | 58.7 | 54 | 54.6 | 50.3 | 54.2 | |
| LwFBackbone=VGG-162021.03 | 45.1 | 45.5 | 53.5 | 57.6 | 56.2 | 65.7 | 63.5 | 58.4 | 59.6 | 58.5 | 56.4 | |
| FinetuningBackbone=VGG-162021.03 | 38.1 | 36 | 43.2 | 44.1 | 45.5 | 54.5 | 50.3 | 50.5 | 51 | 61.2 | 47.4 | |
| EWCBackbone=VGG-162021.03 | 33.9 | 35.1 | 43.6 | 46.7 | 49.5 | 52.5 | 47.8 | 50.2 | 56.6 | 61.4 | 47.8 | |
| LfLBackbone=VGG-162021.03 | 32.4 | 35.1 | 43.4 | 44.1 | 45 | 55.9 | 49.4 | 51.1 | 58.6 | 61.4 | 47.7 |