Image Reconstruction on CelebA
0.995SSIMMOL
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| MOLLearning Paradigm=MOL, Step=Best, Specific Step=40962025.04 | 0.995 | — | — | |
| MOLLearning Paradigm=MOL, Step=40962025.04 | 0.995 | — | — | |
| OursLearning Paradigm=MCL, Step=Best, Specific Step=40962025.04 | 0.995 | — | — | |
| OursLearning Paradigm=MCL, Step=40962025.04 | 0.995 | — | — | |
| OLLearning Paradigm=OL, Step=Best, Specific Step=40962025.04 | 0.969 | — | — | |
| OLLearning Paradigm=OL, Step=40962025.04 | 0.969 | — | — | |
| MAML+CLLearning Paradigm=MCL, Step=Best, Specific Step=642025.04 | 0.92 | — | — | |
| OMLLearning Paradigm=MCL, Step=Best, Specific Step=322025.04 | 0.884 | — | — | |
| MAML+CLLearning Paradigm=MCL, Step=12025.04 | 0.822 | — | — | |
| MOLLearning Paradigm=MOL, Step=12025.04 | 0.82 | — | — | |
| OursLearning Paradigm=MCL, Step=12025.04 | 0.817 | — | — | |
| MAML+CLLearning Paradigm=MCL, Step=40962025.04 | 0.816 | — | — | |
| OMLLearning Paradigm=MCL, Step=12025.04 | 0.796 | — | — | |
| MERLearning Paradigm=CML, Step=Best, Specific Step=2562025.04 | 0.711 | — | — | |
| OMLLearning Paradigm=MCL, Step=40962025.04 | 0.65 | — | — | |
| MERLearning Paradigm=CML, Step=40962025.04 | 0.633 | — | — | |
| EWCLearning Paradigm=CL, Step=Best, Specific Step=2562025.04 | 0.63 | — | — | |
| ERLearning Paradigm=CL, Step=Best, Specific Step=642025.04 | 0.547 | — | — | |
| CLLearning Paradigm=CL, Step=Best, Specific Step=642025.04 | 0.519 | — | — | |
| ERLearning Paradigm=CL, Step=40962025.04 | 0.504 | — | — | |
| EWCLearning Paradigm=CL, Step=40962025.04 | 0.503 | — | — | |
| CLLearning Paradigm=CL, Step=40962025.04 | 0.463 | — | — | |
| MERLearning Paradigm=CML, Step=12025.04 | 0.332 | — | — | |
| OLLearning Paradigm=OL, Step=12025.04 | 0.329 | — | — | |
| EWCLearning Paradigm=CL, Step=12025.04 | 0.321 | — | — | |
| ERLearning Paradigm=CL, Step=12025.04 | 0.316 | — | — | |
| CLLearning Paradigm=CL, Step=12025.04 | 0.313 | — | — | |
| C-ViViTNumber of frames=8-frame2026.03 | — | 5.72 | 24.66 | |
| C-ViViTNumber of frames=18-frame2026.03 | — | 2.27 | 29.68 | |
| LlamaGenDownsample ratio=16, Resolution=512x512, Evaluation Protocol=Zero-shot2025.07 | — | 2.45 | 26.09 | |
| MGVQ-G4Downsample ratio=16, Resolution=512x512, Evaluation Protocol=Zero-shot, Sub-codebooks=42025.07 | — | 0.69 | 31.25 | |
| OpenMagvit2Downsample ratio=16, Resolution=512x512, Evaluation Protocol=Zero-shot2025.07 | — | 1.35 | 27.41 | |
| SD-VAEDownsample ratio=16, Resolution=512x512, Evaluation Protocol=Zero-shot, Tokenizer Type=Continuous2025.07 | — | 0.73 | 30.81 | |
| VARDownsample ratio=16, Resolution=512x512, Evaluation Protocol=Zero-shot2025.07 | — | 1.73 | 26.5 | |
| VQGANDownsample ratio=16, Resolution=512x512, Evaluation Protocol=Zero-shot2025.07 | — | 7.19 | 25.43 | |
| VQVAENumber of frames=1-frame2026.03 | — | 24.62 | 20.52 |