Research map / 3D vision / NeRF and radiance fields
Scene Representation Transformer: Geometry-Free Novel View Synthesis Through Set-Latent Scene Representations
Check 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 a set-latent scene representation and ray-query decoder | SRT · NMR | PSNR: 26.80 -> 27.87 | Helped |
| use a set-latent scene representation and ray-query decoder | SRT · MultiShapeNet | PSNR: 21.97 -> 23.41 | Helped |
| use a set-latent scene representation and ray-query decoder | SRT · MultiShapeNet | LPIPS: 0.332 -> 0.369 | Hurt |
| use a set-latent scene representation and ray-query decoder | SRT · NMR | encoding time: 0.005 s -> 0.010 s | Mixed |
| use a set-latent scene representation and ray-query decoder | SRT · NMR new-scene video rendering | time to encode scene and render video: 75.5 s -> 0.182 s | Helped |
| add an encoder transformer before the decoder (ablation) | SRT · MultiShapeNet | PSNR: 21.64 -> 23.41 | Helped |
| use a set-latent representation and decoder transformer instead of a flat latent with 8-layer MLP (ablation) | SRT · MultiShapeNet | PSNR: 20.88 -> 23.41 | Helped |
| omit test-time camera poses (ablation) | SRT · MultiShapeNet, 5 input images | PSNR: 19.2 -> 22.2 | Helped |
| increase LFN latent dimension from 256 to 1024 (ablation) | LFN · MultiShapeNet novel scenes | PSNR | No clear effect |
| replace ray-based light-field decoding with volumetric rendering (ablation) | SRT · MultiShapeNet | rendering speed: 192× | Hurt |
Closest papers by meaning
- LVSM: A Large View Synthesis Model with Minimal 3D Inductive Bias (ICLR 2025 · similarity 0.88)
- Less Decoder is More Encoder: Geometric Representation Learning from Novel View Synthesis (NeurIPS 2026 · similarity 0.88)
- GoLF-NRT: Integrating Global Context and Local Geometry for Few-Shot View Synthesis (CVPR 2025 · similarity 0.87)
- ERUPT: Efficient Rendering with Unposed Patch Transformer (CVPR 2025 · similarity 0.87)
- Know Your Neighbors: Improving Single-View Reconstruction via Spatial Vision-Language Reasoning (CVPR 2024 · similarity 0.87)
- NViST: In the Wild New View Synthesis from a Single Image with Transformers (CVPR 2024 · similarity 0.87)
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.