Research map / 3D vision / Depth and stereo
RayRoPE: Projective Ray Positional Encoding for Multi-view Attention
Yu Wu, Minsik Jeon, Rick Chang, Oncel Tuzel, Shubham Tulsiani
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 |
|---|---|---|---|
| replace PRoPE with RayRoPE | LVSM · CO3D NVS | PSNR: 17.49 -> 18.40 | Improved |
| replace PRoPE with RayRoPE | LVSM · RE10K NVS | PSNR: 24.48 -> 26.07 | Improved |
| scale to large model, RayRoPE over PRoPE | LVSM (Large) · CO3D NVS | PSNR: 19.77 -> 20.23 | Improved |
| add uncertainty prediction (expected RoPE) (ablation) | RayRoPE · Objaverse NVS | PSNR: 20.37 -> 22.42 | Improved |
| add uncertainty prediction (expected RoPE) (ablation) | RayRoPE · CO3D NVS | PSNR: 17.28 -> 18.40 | Improved |
| add depth prediction (geometric adaptiveness) (ablation) | RayRoPE · CO3D NVS | PSNR: 17.06 -> 18.40 | Improved |
| add depth prediction (geometric adaptiveness) (ablation) | RayRoPE · Objaverse NVS | PSNR: 22.42 -> 22.42 | No clear effect |
| add multi-frequency encoding (ablation) | RayRoPE · RE10K NVS | PSNR: 24.27 -> 26.07 | Improved |
| apply encoding to value and output features (ablation) | RayRoPE · CO3D NVS | PSNR: 18.00 -> 18.40 | Improved |
| supply ground-truth depth to RayRoPE | RayRoPE · CO3D NVS | PSNR: 18.40 -> 20.47 | Improved |
| supply ground-truth depth to PRoPE | PRoPE · CO3D NVS | PSNR: 17.49 -> 19.10 | Improved |
| apply RayRoPE to UniMatch instead of PRoPE | UniMatch · ScanNet stereo depth (unseen) | Abs Rel: 0.101 -> 0.095 | Improved |
Later papers that compared with this method
1 later accepted paper reports RayRoPE as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- URoPE: Universal Relative Position Embedding across Geometric Spaces (ECCV 2026 · better on 2 of 2 shared results)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Cameras as Relative Positional Encoding (NeurIPS 2025 · similarity 0.91)
- GTA: A Geometry-Aware Attention Mechanism for Multi-View Transformers (ICLR 2024 · similarity 0.88)
- DPPE: Rethinking Camera-Based Positional Encoding for Scaling Multi-View Transformers (NeurIPS 2026 · similarity 0.87)
- CRePE: Curved Ray Expectation Positional Encoding for Unified-Camera-Controlled Video Generation (NeurIPS 2026 · similarity 0.87)
- A Global Depth-Range-Free Multi-View Stereo Transformer Network with Pose Embedding (NeurIPS 2024 · similarity 0.86)
- Less Decoder is More Encoder: Geometric Representation Learning from Novel View Synthesis (NeurIPS 2026 · similarity 0.86)
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.
