Efficient View Synthesis and 3D-based Multi-Frame Denoising with Multiplane Feature Representations
About
While current multi-frame restoration methods combine information from multiple input images using 2D alignment techniques, recent advances in novel view synthesis are paving the way for a new paradigm relying on volumetric scene representations. In this work, we introduce the first 3D-based multi-frame denoising method that significantly outperforms its 2D-based counterparts with lower computational requirements. Our method extends the multiplane image (MPI) framework for novel view synthesis by introducing a learnable encoder-renderer pair manipulating multiplane representations in feature space. The encoder fuses information across views and operates in a depth-wise manner while the renderer fuses information across depths and operates in a view-wise manner. The two modules are trained end-to-end and learn to separate depths in an unsupervised way, giving rise to Multiplane Feature (MPF) representations. Experiments on the Spaces and Real Forward-Facing datasets as well as on raw burst data validate our approach for view synthesis, multi-frame denoising, and view synthesis under noisy conditions.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Denoising | Spaces frame 6 (test) | PSNR38 | 48 | |
| Denoising | Spaces (val) | PSNR37.6 | 40 | |
| Novel View Synthesis | Spaces (test) | PSNR35.73 | 24 | |
| Denoising | LLFF-N | PSNR (Gain 1)38.06 | 4 | |
| View Synthesis | Spaces dense 12 views (val) | PSNR35.73 | 4 | |
| View Synthesis | Spaces small 4 views (val) | PSNR33.2 | 4 | |
| View Synthesis | Spaces medium 4 views (val) | PSNR33.47 | 4 | |
| View Synthesis | Spaces large 4 views (val) | PSNR32.38 | 4 | |
| Novel View Synthesis | LLFF-N | Gain 1 PSNR24.52 | 4 |