Video Reconstruction on WebVid 10M
35.76PSNRWF-VAE-L
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| WF-VAE-LTCPR=256(4 x 8 x 8), Chn=162024.11 | 35.76 | 0.943 | 0.023 | 54.36 | — | |
| CogVideoX-VAETCPR=256(4 x 8 x 8), Chn=162024.11 | 35.72 | 0.9434 | 0.0277 | 59.83 | — | |
| PyraTokParams=192M, Latency (ms)=4922026.01 | 35.72 | 0.879 | 0.066 | — | — | |
| LC-VAETCPR=256(4 × 8 × 8), Chn.=162026.04 | 35.59 | 0.9439 | 0.0152 | 73.66 | — | |
| WF-VAETCPR=256(4 × 8 × 8), Chn.=162026.04 | 34.62 | 0.9301 | 0.0193 | 68.72 | — | |
| Hunyuan VAEDownsample Factor=8 × 8 × 4, Total Compression=1:482026.02 | 33.99 | 0.923 | 0.031 | — | 63.72 | |
| LC-VAETCPR=256(4 × 8 × 8), Chn.=82026.04 | 33.78 | 0.9208 | 0.0211 | 135.99 | — | |
| CogVideoX-1.5 VAEDownsample Factor=8 × 8 × 4, Total Compression=1:482026.02 | 33.53 | 0.92 | 0.043 | — | 80.98 | |
| Wan-2.1 VAEDownsample Factor=8 × 8 × 4, Total Compression=1:482026.02 | 33.12 | 0.913 | 0.032 | — | 71.83 | |
| LARPParams=183M, Latency (ms)=6892026.01 | 33.03 | 0.851 | 0.091 | — | — | |
| 3D-MBQ-VAEParams=317M, Latency (ms)=6502026.01 | 33 | 0.848 | 0.092 | — | — | |
| WAN 2.2Params=222M, Latency (ms)=4492026.01 | 32.94 | 0.841 | 0.101 | — | — | |
| WF-VAE-LTCPR=256(4 x 8 x 8), Chn=42024.11 | 32.32 | 0.892 | 0.0513 | 186 | — | |
| SweetTokParams=128M, Latency (ms)=4322026.01 | 32.32 | 0.842 | 0.137 | — | — | |
| AllegroTCPR=256(4 x 8 x 8), Chn=42024.11 | 32.18 | 0.8963 | 0.0524 | 209.68 | — | |
| OmniTokenizerParams=82M, Latency (ms)=4442026.01 | 32.03 | 0.812 | 0.152 | — | — | |
| WF-VAETCPR=256(4 × 8 × 8), Chn.=82026.04 | 31.96 | 0.9281 | 0.0242 | 101.06 | — | |
| Step-Video VAEDownsample Factor=16 × 16 × 8, Total Compression=1:962026.02 | 31.86 | 0.894 | 0.05 | — | 96.75 | |
| VidTokDownsample Factor=8 × 8 × 4, Total Compression=1:962026.02 | 31.79 | 0.887 | 0.048 | — | 124.46 | |
| LC-VAETCPR=256(4 × 8 × 8), Chn.=42026.04 | 31.49 | 0.9207 | 0.0249 | 165.88 | — | |
| WF-VAE-STCPR=256(4 x 8 x 8), Chn=42024.11 | 31.39 | 0.8737 | 0.0517 | 188.04 | — | |
| TokLIPParams=207M, Latency (ms)=6042026.01 | 31.28 | 0.837 | 0.152 | — | — | |
| SVD-VAETCPR=64(1 x 8 x 8), Chn=42024.11 | 31.18 | 0.8689 | 0.0546 | 188.74 | — | |
| SVD-VAETCPR=64(1 × 8 × 8), Chn.=42026.04 | 31.18 | 0.8689 | 0.0546 | 188.74 | — | |
| Open-Sora VAETCPR=256(4 x 8 x 8), Chn=42024.11 | 31.14 | 0.8572 | 0.1001 | 475.23 | — | |
| Open-Sora VAETCPR=256(4 × 8 × 8), Chn.=42026.04 | 31.14 | 0.8572 | 0.1001 | 475.23 | — | |
| FSAE-StandardDownsample Factor=64 × 64 × 4, Total Compression=1:3842026.02 | 30.91 | 0.872 | 0.058 | — | 203.19 | |
| LTX-VideoDownsample Factor=32 × 32 × 8, Total Compression=1:1922026.02 | 30.89 | 0.868 | 0.077 | — | 232.02 | |
| CV-VAETCPR=256(4 x 8 x 8), Chn=42024.11 | 30.76 | 0.8566 | 0.0803 | 369.23 | — | |
| Cosmos-CVDownsample Factor=8 × 8 × 8, Total Compression=1:962026.02 | 30.76 | 0.864 | 0.113 | — | 211.33 | |
| CV-VAETCPR=256(4 × 8 × 8), Chn.=42026.04 | 30.76 | 0.8566 | 0.0803 | 369.23 | — | |
| OD-VAETCPR=256(4 x 8 x 8), Chn=42024.11 | 30.69 | 0.8635 | 0.0553 | 255.92 | — | |
| OD-VAETCPR=256(4 × 8 × 8), Chn.=42026.04 | 30.69 | 0.8635 | 0.0553 | 255.92 | — | |
| WF-VAETCPR=256(4 × 8 × 8), Chn.=42026.04 | 30.68 | 0.9071 | 0.0344 | 179.13 | — | |
| FSAE-LiteDownsample Factor=64 × 64 × 4, Total Compression=1:3842026.02 | 30.42 | 0.861 | 0.075 | — | 240.03 | |
| LG-VQParams=168M, Latency (ms)=5982026.01 | 30.23 | 0.807 | 0.173 | — | — | |
| SD-VAETCPR=64(1 x 8 x 8), Chn=42024.11 | 30.19 | 0.8377 | 0.0568 | 284.9 | — | |
| SD-VAETCPR=64(1 × 8 × 8), Chn.=42026.04 | 30.19 | 0.8377 | 0.0568 | 284.9 | — | |
| CogVideoXParams=288M, Latency (ms)=7122026.01 | 29.92 | 0.811 | 0.141 | — | — | |
| VidTokDownsample Factor=8 × 8 × 4, Total Compression=1:1922026.02 | 29.85 | 0.843 | 0.077 | — | 217.77 | |
| VideoVae+Params=192M, Latency (ms)=5552026.01 | 29.17 | 0.812 | 0.201 | — | — | |
| TokenFlowParams=176M, Latency (ms)=6002026.01 | 28.21 | 0.799 | 0.189 | — | — | |
| Cosmos-CVDownsample Factor=16 × 16 × 8, Total Compression=1:3842026.02 | 28.09 | 0.786 | 0.18 | — | 518.86 | |
| TexTokParams=173M, Latency (ms)=6612026.01 | 27.42 | 0.831 | 0.222 | — | — | |
| VidTokDownsample Factor=16 × 16 × 4, Total Compression=1:7682026.02 | 25.56 | 0.713 | 0.138 | — | 625.32 |