Research map / 3D vision / Depth and stereo
Efficient and High-Quality Depth Estimation via Pixel-Space Diffusion with Linear Attention
Bingde Liu, Wu Ran, Jinglei Zhang, Huanhuan Yuan, Chao Ma
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 |
|---|---|---|---|
| add Global PCM (DINOv2 semantic alignment) (ablation) | Lapis · zero-shot depth avg | AbsRel: 19.8 -> 4.8 | Improved |
| add Local PCM (ablation) | Lapis · zero-shot depth avg | AbsRel: 5.1 -> 4.8 | Improved |
| add Pixel-level Refinement Module (ablation) | Lapis · zero-shot depth avg | AbsRel: 4.9 -> 4.8 | No clear effect |
| use ELU+1 linear attention instead of softmax (ablation) | Lapis · zero-shot depth avg | AbsRel: 5.0 -> 4.8 | No clear effect |
| use AdaLN/MLP conditioning instead of linear cross-attention (ablation) | Lapis · zero-shot depth avg | AbsRel: 10.8 -> 4.8 | Improved |
| use x-prediction instead of v-prediction (ablation) | Lapis · zero-shot depth | AbsRel | Improved |
| use more sampling steps than one (ablation) | Lapis · zero-shot depth | AbsRel | Hurt |
| use Lapis instead of Pixel-Perfect Depth at 1080P | Lapis · 1080P zero-shot depth avg | latency (ms): 905.5 -> 119.0 | Improved |
| use Lapis vs Pixel-Perfect Depth at native resolution | Lapis · native zero-shot depth avg | AbsRel: 4.2 -> 4.1 | Improved |
| use Lapis vs MoGe-2 for speed at 1080P | Lapis · 1080P runtime | latency (ms): 115.3 -> 119.0 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- MoGe-2 — MoGe-2: Accurate Monocular Geometry with Metric Scale and Sharp Details (NeurIPS 2025) · better on 9 of 10 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
- Pixel-Perfect Depth with Semantics-Prompted Diffusion Transformers (NeurIPS 2025 · similarity 0.91)
- DepthFM: Fast Generative Monocular Depth Estimation with Flow Matching (AAAI 2025 · similarity 0.88)
- SharpDepth: Sharpening Metric Depth Predictions Using Diffusion Distillation (CVPR 2025 · similarity 0.88)
- Repurposing Diffusion-Based Image Generators for Monocular Depth Estimation (CVPR 2024 · similarity 0.88)
- More Than Generation: Unifying Generation and Depth Estimation via Text-to-Image Diffusion Models (NeurIPS 2025 · similarity 0.88)
- Novel View Synthesis with Pixel-Space Diffusion Models (CVPR 2025 · 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.
