| Full name | Learnable Infinite Taylor Gaussian for Dynamic View Rendering |
|---|---|
| Publication title | Learnable Infinite Taylor Gaussian for Dynamic View Rendering |
| Publication authors | Bingbing Hu1, Yanyan Li, Rui Xie, Bo Xu, Haoye Dong, Junfeng Yao, Gim Hee Lee |
| Publication venue | CVPR 2025 |
| Publication url | https://arxiv.org/pdf/2412.04282v2 |
| Project link | https://ellisonking.github.io/TaylorGaussian/ |
| Hardware | RTX3080 x 1 |
| Code | Code available with license: MIT License |
| Use other data | No |
| Last edited | 23 May, 2025 |
| PSNR | SSIM | LPIPS | JOD | |
|---|---|---|---|---|
| Left camera | 25.3471 | 0.8430 | 0.3450 | 6.7631 |
| Center camera | 25.9447 | 0.8367 | 0.3667 | 6.7094 |
| Right camera | 25.6876 | 0.8567 | 0.3089 | 6.8287 |
| Total | 25.6598 | 0.8455 | 0.3402 | 6.7671 |
| Subject ID | Left camera | Center camera | Right camera | Ground truth |
|---|---|---|---|---|
| 388 |
|
|
|
|
| 422 |
|
|
|
|
| 443 |
|
|
|
|
| 445 |
|
|
|
|
| 475 |
|
|
|
|