Research map / 3D vision / Gaussian splatting
Mixed Neural Voxels for Fast Multi-view Video Synthesis
Feng Wang, Sinan Tan, Xinghang Li, Zeyue Tian, Yafei Song, Huaping Liu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| split static and dynamic voxels with the variation field instead of a full-dynamic model (ablation) | MixVoxels · coffee-martini | PSNR: 28.36 -> 29.47 | Improved |
| split static and dynamic voxels (ablation) | MixVoxels · coffee-martini | training time: 2.5h -> 0.6h | Improved |
| use inner product time query instead of concatenation (ablation) | MixVoxels · coffee-martini | PSNR: 28.95 -> 29.47 | Improved |
| use inner product time query instead of concatenation (ablation) | MixVoxels · coffee-martini | training time: 58m -> 40m | Improved |
| use inner product time query instead of Fourier head (ablation) | MixVoxels · coffee-martini | PSNR: 28.67 -> 29.47 | Improved |
| increase number of time queries per iteration (ablation) | MixVoxels · coffee-martini | visual detail of moving parts | Improved |
| increase training schedule from MixVoxels-S to MixVoxels-X (ablation) | MixVoxels · Plenoptic Video Dataset | PSNR: 31.03 -> 31.73 | Improved |
| weight variation field by transmittance to handle occlusion (ablation) | MixVoxels variation field · Plenoptic Video Dataset | efficiency and quality | No clear effect |
| MixVoxels-S versus DyNeRF training | DyNeRF · Plenoptic Video Dataset | PSNR: 29.58 -> 31.03 | Improved |
| increase sampling points 8x and iterations in MixVoxels-X | MixVoxels · Plenoptic Video Dataset | render speed (FPS): 37.7 -> 4.6 | Hurt |
Closest papers by meaning
- DynaVol: Unsupervised Learning for Dynamic Scenes through Object-Centric Voxelization (ICLR 2024 · similarity 0.85)
- DiVa-360: The Dynamic Visual Dataset for Immersive Neural Fields (CVPR 2024 · similarity 0.85)
- Learning with Unreliability: Fast Few-shot Voxel Radiance Fields with Relative Geometric Consistency (CVPR 2024 · similarity 0.84)
- Sparse Voxels Rasterization: Real-time High-fidelity Radiance Field Rendering (CVPR 2025 · similarity 0.84)
- 4D3R: Motion-Aware Neural Reconstruction and Rendering of Dynamic Scenes from Monocular Videos (NeurIPS 2025 · similarity 0.84)
- MultimodalStudio: A Heterogeneous Sensor Dataset and Framework for Neural Rendering across Multiple Imaging Modalities (CVPR 2025 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
