Research map / 3D vision / NeRF and radiance fields
HexPlane: A Fast Representation for Dynamic Scenes
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 |
|---|---|---|---|
| replace DyNeRF with HexPlane | dynamic novel view synthesis · Plenoptic Video (Flame Salmon) | Training Time: 1344h -> 12h | Helped |
| replace DyNeRF with HexPlane at the same iteration count | novel view synthesis · Plenoptic Video (Flame Salmon) | LPIPS: 0.099 -> 0.078 | Helped |
| train HexPlane for 100k rather than 650k iterations (ablation) | HexPlane · Plenoptic Video (Flame Salmon) | Training Time: 12h -> 2h | Helped |
| use HexPlane instead of TiNeuVox-S | monocular dynamic scene reconstruction · D-NeRF | Training Time: 12m 10s -> 11m 30s | Helped |
| use HexPlane instead of TiNeuVox-B | monocular dynamic scene reconstruction · D-NeRF | LPIPS: 0.04 -> 0.04 | No clear effect |
| use spatial-temporal planes rather than Volume Basis (ablation) | dynamic scene feature representation · D-NeRF | PSNR: 30.631 -> 31.042 | Helped |
| use spatial-temporal planes rather than VM-T (ablation) | dynamic scene feature representation · D-NeRF | PSNR: 30.744 -> 31.042 | Helped |
| use spatial-temporal planes rather than CP Decomposition (ablation) | dynamic scene feature representation · D-NeRF | PSNR: 30.302 -> 31.042 | Helped |
| add spatial-temporal planes to purely spatial planes (ablation) | HexPlane · D-NeRF | PSNR: 20.369 -> 31.042 | Helped |
| add purely spatial planes to spatial-temporal planes (ablation) | HexPlane · D-NeRF | PSNR: 21.112 -> 31.042 | Helped |
| use all complementary plane pairs rather than one XY-ZT pair (ablation) | HexPlane · D-NeRF | PSNR: 30.370 -> 31.042 | Helped |
| use complementary plane axes rather than repeated-axis pairings (ablation) | HexPlane · D-NeRF | PSNR: 28.562 -> 31.042 | Helped |
Closest papers by meaning
- Real-time Photorealistic Dynamic Scene Representation and Rendering with 4D Gaussian Splatting (ICLR 2024 · similarity 0.87)
- DynamicCity: Large-Scale 4D Occupancy Generation from Dynamic Scenes (ICLR 2025 · similarity 0.87)
- EmerNeRF: Emergent Spatial-Temporal Scene Decomposition via Self-Supervision (ICLR 2024 · similarity 0.85)
- Flux4D: Flow-based Unsupervised 4D Reconstruction (NeurIPS 2025 · similarity 0.85)
- DaReNeRF: Direction-aware Representation for Dynamic Scenes (CVPR 2024 · similarity 0.85)
- 4D3R: Motion-Aware Neural Reconstruction and Rendering of Dynamic Scenes from Monocular Videos (NeurIPS 2025 · similarity 0.85)
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.