Research map / 3D vision / 3D generation
ShapeR: Robust Conditional 3D Shape Generation from Casual Captures
Yawar Siddiqui, Duncan Frost, Samir Aroudj, Armen Avetisyan, Henry Howard-Jenkins, Daniel DeTone, Pierre Moulon, Qirui Wu, Zhengqin Li, Julian Straub, Richard Newcombe, Jakob Engel
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 |
|---|---|---|---|
| add SLAM points as an input modality (ablation) | ShapeR · ShapeR Evaluation Dataset | Chamfer distance x10^2: 4.514 -> 2.375 | Improved |
| add point augmentations (ablation) | ShapeR · ShapeR Evaluation Dataset | Chamfer distance x10^2: 3.276 -> 2.375 | Improved |
| add image augmentations (ablation) | ShapeR · ShapeR Evaluation Dataset | Chamfer distance x10^2: 3.397 -> 2.375 | Improved |
| add scene-level second-stage fine-tuning (ablation) | ShapeR · ShapeR Evaluation Dataset | Chamfer distance x10^2: 3.053 -> 2.375 | Improved |
| add 2D point mask prompting of DINO features (ablation) | ShapeR · ShapeR Evaluation Dataset | Chamfer distance x10^2: 2.568 -> 2.375 | Improved |
| add 2D point mask prompting of DINO features (ablation) | ShapeR · ShapeR Evaluation Dataset | Normal consistency: 0.813 -> 0.810 | No clear effect |
| add point augmentations (ablation) | ShapeR · ShapeR Evaluation Dataset | Normal consistency: 0.805 -> 0.810 | No clear effect |
| add SLAM points (ablation) | ShapeR · ShapeR Evaluation Dataset | F1: 0.486 -> 0.722 | Improved |
| use ShapeR instead of best posed multi-view baseline (FStereo) | FStereo · ShapeR Evaluation Dataset | Chamfer distance x10^2: 6.483 -> 2.375 | Improved |
| use ShapeR instead of image-to-3D foundation models with manual views and SAM2 masks | TripoSG/Direct3DS2/Hunyuan3D-2.0/Amodal3R · User study | user preference: 81.11% to 88.33% win rate | Improved |
Closest papers by meaning
- High-Fidelity 3D Textured Shapes Generation by Sparse Encoding and Adversarial Decoding (ECCV 2024 · similarity 0.88)
- 3DMiner: Discovering Shapes from Large-Scale Unannotated Image Datasets (ICCV 2023 · similarity 0.87)
- LASR: Learning Articulated Shape Reconstruction From a Monocular Video (CVPR 2021 · similarity 0.86)
- ShapeFormer: Transformer-Based Shape Completion via Sparse Representation (CVPR 2022 · similarity 0.86)
- SDFit: 3D Object Pose and Shape by Fitting a Morphable SDF to a Single Image (ICCV 2025 · similarity 0.86)
- ArtFormer: Controllable Generation of Diverse 3D Articulated Objects (CVPR 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.
