Research map / 3D vision / Face and shape
Learning View Priors for Single-View 3D Reconstruction
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 |
|---|---|---|---|
| Add view prior learning (ablation) | ShapeNet single-view reconstruction · ShapeNet single-view | IoU: .403 -> .490 | Helped |
| Add class conditioning to the VPL discriminator (ablation) | ShapeNet single-view reconstruction · ShapeNet single-view | IoU: .490 -> .505 | Helped |
| Add texture prediction to VPL with class conditioning (ablation) | ShapeNet single-view reconstruction · ShapeNet single-view | IoU: .505 -> .513 | Helped |
| Use VPL with class conditioning (ablation) | ShapeNet multi-view reconstruction · ShapeNet multi-view, 20 views | IoU: .652 -> .655 | Helped |
| Use VPL (ablation) | PASCAL category-agnostic reconstruction · PASCAL single-view, category-agnostic | IoU: .454 -> .473 | Helped |
| Use VPL (ablation) | PASCAL category-specific reconstruction · PASCAL single-view, silhouette-only | IoU: .472 -> .488 | Helped |
| Add internal pressure (ablation) | ShapeNet single-view reconstruction · ShapeNet single-view | IoU: .387 -> .403 | Helped |
| Add symmetricity to internal pressure (ablation) | ShapeNet single-view reconstruction · ShapeNet single-view | IoU: .403 -> .420 | Helped |
| Add graph Laplacian regularization (ablation) | ShapeNet single-view reconstruction with internal pressure · ShapeNet single-view | IoU: .403 -> .403 | No clear effect |
| Add edge-length regularization (ablation) | ShapeNet single-view reconstruction with internal pressure · ShapeNet single-view | IoU: .403 -> .403 | No clear effect |
| Replace the estimated-view discriminator with a discriminator of reconstructed views versus real dataset views (ablation) | ShapeNet single-view reconstruction · ShapeNet single-view | IoU: .403 -> .403 | No clear effect |
| Replace gradient reversal with iterative optimization for the estimated-view discriminator (ablation) | ShapeNet single-view reconstruction · ShapeNet single-view | IoU: .505 -> .514 | No clear effect |
Closest papers by meaning
- Fostering Generalization in Single-View 3D Reconstruction by Learning a Hierarchy of Local and Global Shape Priors (CVPR 2021 · similarity 0.91)
- Novel Object Viewpoint Estimation Through Reconstruction Alignment (CVPR 2020 · similarity 0.90)
- The More You See in 2D the More You Perceive in 3D (CVPR 2024 · similarity 0.88)
- Learning Partonomic 3D Reconstruction from Image Collections (CVPR 2025 · similarity 0.87)
- NOPE: Novel Object Pose Estimation from a Single Image (CVPR 2024 · similarity 0.87)
- Decompositional Neural Scene Reconstruction with Generative Diffusion Prior (CVPR 2025 · 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.