Text Based Image Editing On Pie Bench
Metrics
Background LPIPS
Background PSNR
CLIPSIM
Structure Distance
Results
Performance results of various models on this benchmark
Comparison Table
| Model Name | Background LPIPS | Background PSNR | CLIPSIM | Structure Distance |
|---|---|---|---|---|
| direct-inversion-boosting-diffusion-based | 87.94 | 22.64 | 24.38 | 24.70 |
| direct-inversion-boosting-diffusion-based | 138.98 | 21.53 | 23.31 | 49.22 |
| fireflow-fast-inversion-of-rectified-flow-for | 123.6 | 23.03 | 26.02 | 27.1 |
| prompt-to-prompt-image-editing-with-cross | 208.80 | 17.87 | 25.01 | 69.43 |
| negative-prompt-inversion-fast-image | 69.01 | 26.21 | 24.61 | 16.17 |
| masactrl-tuning-free-mutual-self-attention | 106.62 | 22.17 | 23.96 | 28.38 |
| inversion-free-image-editing-with-natural | 55.85 | 26.64 | 24.57 | 15.61 |
| stylediffusion-prompt-embedding-inversion-for | 66.10 | 26.05 | 24.78 | 11.65 |
| inversion-free-image-editing-with-natural | 47.98 | 27.52 | 24.89 | 14.22 |
| zero-shot-image-to-image-translation | 172.22 | 20.44 | 22.80 | 61.68 |
| direct-inversion-boosting-diffusion-based | 54.55 | 27.22 | 25.02 | 11.65 |
| plug-and-play-diffusion-features-for-text | 113.46 | 22.28 | 25.41 | 28.22 |
| direct-inversion-boosting-diffusion-based | 106.06 | 22.46 | 25.41 | 24.29 |
| null-text-inversion-for-editing-real-images | 60.67 | 27.03 | 24.75 | 13.44 |
| kv-edit-training-free-image-editing-for | 9.92 | 35.87 | 25.63 | 1.98 |
| vision-guided-and-mask-enhanced-adaptive | 45.67 | 28.27 | 25.91 | 12.65 |
| fireflow-fast-inversion-of-rectified-flow-for | 239.4 | 16.49 | 27.33 | 70.9 |
| inversion-free-image-editing-with-natural | 47.58 | 28.51 | 25.03 | 13.78 |