Research map / 3D vision / Face and shape
DeepSurfels: Learning Online Appearance Fusion
Marko Mihajlovic, Silvan Weder, Marc Pollefeys, Martin R. Oswald
Check 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 |
|---|---|---|---|
| Use deterministic DeepSurfels instead of TSDF Coloring | DeepSurfels · ShapeNet cars | PSNR: 26.68 -> 27.20 | Improved |
| Add depth as a meta feature (ablation) | DeepSurfels learned (32^3, 6x6, 3+3) · ShapeNet cars | PSNR: 28.27 -> 28.31 | Improved |
| Add multi-view consistency regularization (ablation) | DeepSurfels learned (32^3, 6x6, 3+3) · ShapeNet cars | PSNR: 28.31 -> 28.36 | Improved |
| Add viewing direction and surface orientation (ablation) | DeepSurfels learned (32^3, 6x6, 3+3) · ShapeNet cars | PSNR: 28.36 -> 28.89 | Improved |
| Use spatially aligned 6x6 patches instead of 1x1 patches with 213+3 channels (ablation) | DeepSurfels at 32^3 grid resolution · ShapeNet cars | PSNR: 22.95 -> 28.89 | Improved |
| Increase patch resolution from 4x4 to 6x6 (ablation) | DeepSurfels at 64^3 grid resolution · ShapeNet cars | PSNR: 29.92 -> 30.27 | Improved |
| Increase patch resolution from 2x2 to 6x6 (ablation) | DeepSurfels at 128^3 grid resolution · ShapeNet cars | PSNR: 30.23 -> 30.64 | Improved |
| Increase feature channels from 2 to 10 (ablation) | DeepSurfels on the cube dataset · Cube dataset | PSNR: 25.95 -> 28.27 | Improved |
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.
- SurfelMeshing — SurfelMeshing: Online Surfel-Based Mesh Reconstruction (TPAMI 2019) · better on 1 of 1 shared result
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
- Deep Appearance Maps (ICCV 2019 · similarity 0.86)
- Towards Learning Neural Representations from Shadows (ECCV 2022 · similarity 0.85)
- Learning 3D Representations for Spatial Intelligence from Unposed Multi-View Images (CVPR 2026 · similarity 0.85)
- Mesh2Tex: Generating Mesh Textures from Image Queries (ICCV 2023 · similarity 0.85)
- Real-time Free-view Human Rendering from Sparse-view RGB Videos using Double Unprojected Textures (CVPR 2025 · similarity 0.84)
- Neural Implicit Morphing of Face Images (CVPR 2024 · similarity 0.83)
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.
