Affective Visual Customization on L-AVC (test)
0.068FIDEPEM
Evaluation results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| EPEMMLLM=BLIP2 with OPT (2.7B), Generative Model=Stable Diffusion-v1.5, Training=Fine-tuned2026.02 | 0.068 | 0.339 | 58.29 | 82.7 | 27.96 | 59.97 | 75.8 | 80.2 | 9.6 | |
| SmartEditTraining=Fine-tuned2026.02 | 0.098 | 0.397 | 52.08 | 78.89 | 23.31 | 51 | 67.8 | 71.8 | 10.3 | |
| MGIETraining=Fine-tuned2026.02 | 0.099 | 0.39 | 53.56 | 79.18 | 24.92 | 52.82 | 68 | 73 | 9.8 | |
| EPEM w/o EAIVariant=Ablation without EAI block2026.02 | 0.102 | 0.405 | 51.64 | 77.74 | 23.15 | 53.74 | 64.6 | 69.2 | 9.1 | |
| EPEM w/o EICVariant=Ablation without Efficient Inter-emotion Converting2026.02 | 0.107 | 0.435 | 48.31 | 73.26 | 21.93 | 51.75 | 60.4 | 65.4 | 7.5 | |
| SDEditTraining=Fine-tuned2026.02 | 0.115 | 0.441 | 47.67 | 71.08 | 22.45 | 51.08 | 60.2 | 68.8 | 7.1 | |
| DiffEditTraining=Fine-tuned2026.02 | 0.118 | 0.426 | 49.68 | 72.35 | 21.19 | 52.45 | 63.6 | 69 | 7.3 | |
| InstructPix2PixTraining=Fine-tuned2026.02 | 0.125 | 0.508 | 42.71 | 72.76 | 22.56 | 47.42 | 56.6 | 66.2 | 7.2 | |
| Prompt-to-PromptTraining=Fine-tuned2026.02 | 0.131 | 0.558 | 45.64 | 73.27 | 21.57 | 45.22 | 53.4 | 65 | 7.6 | |
| ControlNetTraining=Fine-tuned2026.02 | 0.14 | 0.596 | 29.38 | 69.24 | 20.64 | 38.61 | 42.4 | 58.4 | 8.9 |