Layout-to-Image Generation on COCO-Position 2014
65.97APReal Image
Evaluation results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Real Image2024.02 | 65.97 | 83.75 | 85.49 | 79.11 | 71.22 | 24.22 | 19.74 | — | |
| MIGCBackbone=Stable Diffusion v1.4, Optimizer=AdamW, Scheduler=EulerDiscreteScheduler, Inference Steps=50, MIGC Active Steps=25, CFG Scale=7.52024.02 | 54.69 | 80.29 | 77.38 | 84.17 | 61.71 | 24.66 | 20.25 | 24.52 | |
| GLIGEN2024.02 | 40.68 | 70.52 | 71.61 | 68.26 | 42.85 | 24.61 | 19.69 | 26.8 | |
| ITERINPAINTzero-shot=true, input_format=bounding box, evaluator=YOLOv72023.04 | 39.1 | — | — | — | — | — | — | — | |
| ReCozero-shot=true, input_format=bounding box, evaluator=YOLOv72023.04 | 38.9 | — | — | — | — | — | — | — | |
| GLIGENzero-shot=true, input_format=bounding box, evaluator=YOLOv72023.04 | 36.4 | — | — | — | — | — | — | — | |
| Layout Diffusion2024.02 | 23.45 | 50.53 | 57.49 | 48.1 | 20.7 | 18.28 | 19.08 | 25.94 | |
| ControlNetzero-shot=true, input_format=segmentation mask, evaluator=YOLOv72023.04 | 15.3 | — | — | — | — | — | — | — | |
| Multi Diffusion2024.02 | 6.72 | 23.86 | 38.82 | 18.65 | 3.63 | 22.1 | 19.13 | 33.2 | |
| BOX-Diffusion2024.02 | 3.29 | 17.84 | 33.38 | 12.27 | 1.08 | 23.79 | 18.7 | 25.15 | |
| TFLCG2024.02 | 1.75 | 13.54 | 28.01 | 6.77 | 0.56 | 25.07 | 17.97 | 24.65 | |
| Stable DiffusionBackbone=Stable Diffusion v1.42024.02 | 0.8 | 5.95 | 21.6 | 2.71 | 0.42 | 25.69 | 17.34 | 23.56 |