Research map / 3D vision / Face and shape
Differentiable Volumetric Rendering: Learning Implicit 3D Representations Without 3D Supervision
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 |
|---|---|---|---|
| Train DVR with RGB image supervision instead of DRC mask supervision | DVR · ShapeNet 13 classes, multi-view supervision | Chamfer-L1 (mean): 0.553 -> 0.239 | Helped |
| Train DVR with RGB image supervision instead of SoftRas | DVR · ShapeNet 13 classes, multi-view supervision | Chamfer-L1 (mean): 0.266 -> 0.239 | Helped |
| Train DVR with depth supervision instead of DRC depth supervision | DVR · ShapeNet 13 classes, multi-view depth supervision | Chamfer-L1 (mean): 0.401 -> 0.206 | Helped |
| Use one training view per object rather than 24 views | DVR · ShapeNet single-view reconstruction | Chamfer distance: 0.383 -> 0.410 | Hurt |
| Add sparse depth supervision to RGB supervision (ablation) | DVR · DTU scans 65, 106, and 118 | Chamfer-L1: 0.907 -> 0.782 | Helped |
| Add sparse depth supervision to RGB supervision (ablation) | DVR · DTU scans 65, 106, and 118 | Accuracy: 1.054 -> 0.789 | Helped |
| Add sparse depth supervision to RGB supervision (ablation) | DVR · DTU scans 65, 106, and 118 | Completeness: 0.760 -> 0.775 | Hurt |
Closest papers by meaning
- Differentiable Stereopsis: Meshes From Multiple Views Using Differentiable Rendering (CVPR 2022 · similarity 0.91)
- NeAT: Learning Neural Implicit Surfaces With Arbitrary Topologies From Multi-View Images (CVPR 2023 · similarity 0.90)
- Representing 3D Shapes With Probabilistic Directed Distance Fields (CVPR 2022 · similarity 0.89)
- Learning to Infer Implicit Surfaces without 3D Supervision (NeurIPS 2019 · similarity 0.89)
- LinPrim: Linear Primitives for Differentiable Volumetric Rendering (NeurIPS 2025 · similarity 0.88)
- HISR: Hybrid Implicit Surface Representation for Photorealistic 3D Human Reconstruction (AAAI 2024 · 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.