Research map / 3D vision / 3D generation
3D Assembly Completion
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 FiT's transformer-based part-relation modeling instead of DGL's graph neural network modules | 3D assembly completion · Original toolkit, k=1, Chair | PA (%): 62.13 -> 70.71 | Helped |
| Use FiT's transformer-based part-relation modeling instead of DGL's graph neural network modules | 3D assembly completion · Original toolkit, k=1, Chair | MA (%): 91.05 -> 92.93 | Helped |
| Add pose concatenation at the encoder and memory stages (ablation) | FiT · Original toolkit, k=1, Chair | PA (%): 52.93 -> 70.71 | Helped |
| Add symmetric transformation consistency loss Lsym (ablation) | FiT · Original toolkit, k=1, Chair | PA (%): 66.96 -> 70.71 | Helped |
| Add symmetric transformation consistency loss Lsym (ablation) | FiT · Original toolkit, k=1, Chair | CA (%): 56.31 -> 62.22 | Helped |
| Add translation loss Lt (ablation) | FiT · Original toolkit, k=1, Chair | MA (%): 86.16 -> 92.93 | Helped |
| Add rotation loss Lr (ablation) | FiT · Original toolkit, k=1, Chair | PA (%): 67.56 -> 70.71 | Helped |
| Add shape Chamfer loss Ls (ablation) | FiT · Original toolkit, k=1, Chair | PA (%): 67.95 -> 70.71 | Helped |
| Jointly train candidate matching and pose prediction instead of either task alone (ablation) | FiT · Original toolkit, k=1, Chair | MA (%): 84.67 -> 92.93 | Helped |
| Jointly train candidate matching and pose prediction instead of pose prediction alone (ablation) | FiT · Original toolkit, k=1, Chair | PA (%): 51.05 -> 70.71 | Helped |
| Use 3 encoder and decoder layers rather than 2 (ablation) | FiT · Original toolkit, k=1, Chair | PA (%): 67.99 -> 70.71 | Helped |
| Use 3 encoder and decoder layers rather than 6 (ablation) | FiT · Original toolkit, k=1, Chair | PA (%): 67.42 -> 70.71 | Helped |
Closest papers by meaning
- Neural Assembler: Learning to Generate Fine-Grained Robotic Assembly Instructions from Multi-View Images (AAAI 2025 · similarity 0.88)
- Imagine: Image-Guided 3D Part Assembly with Structure Knowledge Graph (AAAI 2025 · similarity 0.88)
- Two by Two: Learning Multi-Task Pairwise Objects Assembly for Generalizable Robot Manipulation (CVPR 2025 · similarity 0.87)
- IKEA Manuals at Work: 4D Grounding of Assembly Instructions on Internet Videos (NeurIPS 2024 · similarity 0.87)
- Generative 3D Part Assembly via Part-Whole-Hierarchy Message Passing (CVPR 2024 · similarity 0.87)
- Category-Level Multi-Part Multi-Joint 3D Shape Assembly (CVPR 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.