Multi-object semantic discovery on Tetrominoes
99.5mAccDTI-Sprites
Evaluation results
| Method | Links | ||
|---|---|---|---|
| DTI-SpritesSprite generation=Pixels, Decision function=Min-Loss, Activation function=S, Training criterion=L0−1=comp, Regularization=Lempty, Training schedule=Longer training (†)2026.04 | 99.5 | 99.2 | |
| Ours-DSprite generation=MLP, Decision function=linear mapping, Activation function=GS, Training criterion=Lcomp, Regularization=τ + Lfreq2026.04 | 93.9 | 89.9 | |
| Ours-DSprite generation=MLP, Decision function=linear mapping, Activation function=GS, Training criterion=Lcomp, Regularization=τ2026.04 | 92.7 | 89.2 | |
| Ours-DSprite generation=MLP, Decision function=linear mapping, Activation function=GS, Training criterion=Lcomp, Regularization=None2026.04 | 74.3 | 64.4 |