Research map / 3D vision / Depth and stereo
Tensor4D: Efficient Neural 4D Decomposition for High-Fidelity Dynamic Reconstruction and Rendering
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use Tensor4D instead of TiNeuVox | monocular synthetic reconstruction · Lego | PSNR: 25.14 -> 26.71 | Helped |
| Use Tensor4D instead of TiNeuVox | monocular synthetic reconstruction · Standup | PSNR: 36.18 -> 36.32 | Helped |
| Use Tensor4D instead of TiNeuVox | monocular synthetic reconstruction · Jumpingjacks | PSNR: 34.76 -> 34.43 | Hurt |
| Use Tensor4D instead of the strongest listed baseline | sparse-view reconstruction · Sequence1-thz | PSNR: 25.15 -> 28.27 | Helped |
| Use Tensor4D instead of NeuS-T | sparse-view reconstruction · Earphone | PSNR: 24.22 -> 25.00 | Helped |
| Use Tensor4D instead of NeuS-T | sparse-view reconstruction · Sequence3-yxd | PSNR: 26.59 -> 28.83 | Helped |
| Add regularization terms (ablation) | Tensor4D · Lego | PSNR: 26.49 -> 26.71 | Helped |
| Add regularization terms (ablation) | Tensor4D · Standup | PSNR: 35.91 -> 36.32 | Helped |
| Add regularization terms (ablation) | Tensor4D · Sequence-thz | PSNR: 27.92 -> 28.27 | Helped |
| Add regularization terms (ablation) | Tensor4D · Sequence-earphone | PSNR: 24.75 -> 25.00 | Helped |
| Use hierarchical decomposition instead of 6-plane decomposition (ablation) | Tensor4D · Lego | PSNR: 26.44 -> 26.71 | Helped |
| Use hierarchical decomposition instead of 6-plane decomposition (ablation) | Tensor4D · Standup | PSNR: 35.79 -> 36.32 | Helped |
Closest papers by meaning
- Real-time Photorealistic Dynamic Scene Representation and Rendering with 4D Gaussian Splatting (ICLR 2024 · similarity 0.91)
- 4D3R: Motion-Aware Neural Reconstruction and Rendering of Dynamic Scenes from Monocular Videos (NeurIPS 2025 · similarity 0.89)
- 4D-LRM: Large Space-Time Reconstruction Model From and To Any View at Any Time (NeurIPS 2025 · similarity 0.89)
- Flux4D: Flow-based Unsupervised 4D Reconstruction (NeurIPS 2025 · similarity 0.89)
- Swift4D: Adaptive divide-and-conquer Gaussian Splatting for compact and efficient reconstruction of dynamic scene (ICLR 2025 · similarity 0.88)
- Grid4D: 4D Decomposed Hash Encoding for High-Fidelity Dynamic Gaussian Splatting (NeurIPS 2024 · similarity 0.88)
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.