Video Deinterlacing On Msu Deinterlacer
评估指标
FPS on CPU
PSNR
SSIM
Subjective
VMAF
评测结果
各个模型在此基准测试上的表现结果
| Paper Title | Repository | ||||||
|---|---|---|---|---|---|---|---|
| MFDIN (L) | 1.6 | 43.884 | 0.979 | 1.054 | 97.30 | Multi-frame Joint Enhancement for Early Interlaced Videos | - |
| DfRes (SA) | 0.1 | 43.486 | 0.972 | 0.925 | 95.96 | Multi-Field De-interlacing using Deformable Convolution Residual Blocks and Self-Attention | - |
| FLAD | 0.1 | 43.293 | 0.977 | 0.875 | 96.89 | - | - |
| DfRes (122000 G2e 3) | - | 43.200 | 0.972 | 0.862 | 95.68 | Multi-Field De-interlacing using Deformable Convolution Residual Blocks and Self-Attention | - |
| SwinDI | - | 41.151 | 0.970 | 0.741 | 95.17 | - | - |
| EDVR (woTSA) | 0.6 | 41.017 | 0.964 | 0.524 | 94.43 | - | - |
| ST-Deint | 2.7 | 40.869 | 0.964 | 0.550 | 94.36 | Spatial-Temporal Correlation and Topology Learning for Person Re-Identification in Videos | - |
| EDVR | 0.6 | 40.678 | 0.962 | 0.492 | 94.01 | - | - |
| DfRes | 0.4 | 40.590 | 0.971 | 0.912 | 95.20 | Multi-Field De-interlacing using Deformable Convolution Residual Blocks and Self-Attention | - |
| MSU Deinterlacer | 1.3 | 39.846 | 0.966 | 0.733 | 94.32 | - | - |
| DUF | 0.7 | 39.845 | 0.960 | 0.368 | 93.20 | - | - |
| MFDIN | 1.6 | 39.803 | 0.961 | 0.963 | 94.38 | Multi-frame Joint Enhancement for Early Interlaced Videos | - |
| Bob-Weave Deinterlacer | 46.6 | 39.298 | 0.957 | 0.347 | 93.53 | - | - |
| VapourSynth TDeintMod | 50.3 | 38.963 | 0.955 | 0.251 | 93.28 | - | - |
| TDAN | 0.7 | 38.955 | 0.956 | 0.253 | 91.97 | - | - |
| NNEDI | 1.9 | 38.443 | 0.957 | 0.472 | 93.15 | - | - |
| VapourSynth EEDI3 | 51.9 | 38.403 | 0.957 | 0.393 | 92.52 | - | - |
| Real-time Deep Video Deinterlacing | 0.3 | 38.374 | 0.957 | 0.543 | 93.28 | Real-time Deep Video Deinterlacing | |
| YADIF | 49 | 38.260 | 0.949 | 0.621 | 90.13 | - | - |
| Weston 3-Field Deinterlacer | 36.8 | 38.131 | 0.951 | 0.219 | 92.16 | - | - |
0 of 31 row(s) selected.