Research map / 3D vision / Face and shape
Learning to Predict 3D Objects with an Interpolation-based Differentiable Renderer
Read the paperCheck 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 |
|---|---|---|---|
| Use DIB-R instead of SoftR. | Single-image 3D object prediction · ShapeNet 13 categories, mean | 3D IOU (%): 59.3 -> 61.2 | Helped |
| Use DIB-R instead of N3MR | Single-image 3D object prediction · ShapeNet 13 categories, mean | 3D IOU (%): 57.0 -> 61.2 | Helped |
| Use DIB-R instead of SoftR. | Single-image 3D object prediction · ShapeNet 13 categories, mean | F-score (%): 49.9 -> 55.8 | Helped |
| Use DIB-R instead of N3MR | Single-image 3D object prediction · ShapeNet 13 categories, mean | F-score (%): 54.7 -> 55.8 | Helped |
| Use DIB-R instead of N3MR | Texture and lighting prediction · ShapeNet cars | Texture L-1 loss: 0.03640 -> 0.02179 | Helped |
| Use DIB-R instead of N3MR | Texture and lighting prediction · ShapeNet cars | Lighting angle: 23.5585 -> 9.7096 | Helped |
| Use DIB-R instead of N3MR | Texture and lighting prediction · ShapeNet cars | Texture + light L-1 loss: 0.02208 -> 0.01362 | Helped |
| Use DIB-R instead of CMR | Single-view shape and texture prediction · CUB birds | 2D IOU loss: 0.262 -> 0.243 | Helped |
| Use DIB-R instead of CMR | Single-view shape and texture prediction · CUB birds | Key Point: 0.930 -> 0.972 | Helped |
| Use DIB-R instead of CMR | Single-view shape and texture prediction · CUB birds | Texture L-1 loss: 0.043 -> 0.043 | No clear effect |
| Add adversarial loss | DIB-R texture and lighting prediction · ShapeNet cars | Texture crispness | Helped |
| Predict shininess from 2D supervision | DIB-R Phong lighting model · ShapeNet cars | Shininess prediction accuracy | Hurt |
Closest papers by meaning
- DIB-R++: Learning to Predict Lighting and Material with a Hybrid Differentiable Renderer (NeurIPS 2021 · similarity 0.90)
- Neural 3D Mesh Renderer (CVPR 2018 · similarity 0.88)
- Convolutional Generation of Textured 3D Meshes (NeurIPS 2020 · similarity 0.87)
- Enhancing High-Resolution 3D Generation through Pixel-wise Gradient Clipping (ICLR 2024 · similarity 0.86)
- BG-Triangle: Bezier Gaussian Triangle for 3D Vectorization and Rendering (CVPR 2025 · similarity 0.85)
- Diff3DS: Generating View-Consistent 3D Sketch via Differentiable Curve Rendering (ICLR 2025 · similarity 0.85)
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.