Research map / 3D vision / Face and shape
"DECOLLAGE: 3D Detailization by Controllable, Localized, and Learned Geometry Enhancement"
Qimin Chen, Zhiqin Chen, Vladimir G. Kim, Noam Aigerman, Hao Zhang, Siddhartha Chaudhuri
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 |
|---|---|---|---|
| use the pyramid GAN structure instead of single-scale DECOR-GAN design (ablation) | DECOLLAGE multi-category style mixing · Chair+table style mixing | Strict-IOU: 0.621 -> 0.661 | Improved |
| add downsampling structure-preserving loss (ablation) | DECOLLAGE pyramid · Chair+table style mixing | Strict-IOU: 0.661 -> 0.713 | Improved |
| add upsampling structure-preserving loss (ablation) | DECOLLAGE pyramid · Chair+table style mixing | Strict-IOU: 0.661 -> 0.705 | Improved |
| add both L_down and L_up (ablation) | DECOLLAGE pyramid · Chair+table style mixing | Strict-IOU: 0.661 -> 0.753 | Improved |
| add spatially adaptive alpha weighting at style boundaries (ablation) | DECOLLAGE full · Chair+table style mixing | Strict-IOU: 0.753 -> 0.761 | Improved |
| add part labels as input (ablation) | DECOLLAGE · Chair+table style mixing | Cls-score: 0.533 -> 0.518 | Improved |
| replace DECOR-GAN with DECOLLAGE on chair detailization | DECOR-GAN · Chair | Strict-IOU: 0.581 -> 0.748 | Improved |
| DECOLLAGE vs DECOR-GAN on building local patch F-score | DECOLLAGE · Building | LP-F-score: 0.662 -> 0.648 | Hurt |
| DECOLLAGE vs DECOR-GAN on plant local patch IOU | DECOLLAGE · Plant | LP-IOU: 0.385 -> 0.358 | 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.
- DECOR-GAN — DECOR-GAN: 3D Shape Detailization by Conditional Refinement (CVPR 2021) · better on 3 of 5 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
- DECOR-GAN: 3D Shape Detailization by Conditional Refinement (CVPR 2021 · similarity 0.93)
- 3D Paintbrush: Local Stylization of 3D Shapes with Cascaded Score Distillation (CVPR 2024 · similarity 0.86)
- Structured 3D Latents for Scalable and Versatile 3D Generation (CVPR 2025 · similarity 0.86)
- ARMesh: Autoregressive Mesh Generation via Next-Level-of-Detail Prediction (NeurIPS 2025 · similarity 0.86)
- LaGeM: A Large Geometry Model for 3D Representation Learning and Diffusion (ICLR 2025 · similarity 0.86)
- Wukong's 72 Transformations: High-fidelity Textured 3D Morphing via Flow Models (NeurIPS 2025 · 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.
