Research map / Point clouds / Shape and surface completion
KT-Net: Knowledge Transfer for Unpaired 3D Shape Completion
Zhen Cao, Wenxiao Zhang, Xin Wen, Zhen Dong, Yu-Shen Liu, Xiongwu Xiao, Bisheng Yang
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 teacher-assistant-student knowledge transfer | unpaired shape completion · 3D-EPN | CD × 10^4 (Average): 14.3 -> 10.0 | Improved |
| use teacher-assistant-student knowledge transfer | unpaired shape completion · CRN | CD × 10^4 (Average): 14.9 -> 11.5 | Improved |
| add KRA to recover missing complete-shape knowledge (ablation) | KT-Net · CRN (Car, Chair, Lamp, Sofa, Table) | CD × 10^4 (Average): 19.7 -> 13.4 | Improved |
| add KDA to discriminate transferred student features against teacher features (ablation) | KT-Net · CRN (Car, Chair, Lamp, Sofa, Table) | CD × 10^4 (Average): 22.1 -> 13.4 | Improved |
| add residual connection to KRA (ablation) | KT-Net · CRN (Car, Chair, Lamp, Sofa, Table) | CD × 10^4 (Average): 16.2 -> 13.4 | Improved |
| add the student loss constraining the completion to preserve input information (ablation) | KT-Net · CRN (Car, Chair, Lamp, Sofa, Table) | CD × 10^4 (Average): 22.9 -> 13.4 | Improved |
| transfer multi-stage features rather than only the global feature (ablation) | KT-Net · CRN (Car, Chair, Lamp, Sofa, Table) | CD × 10^4 (Average): 19.6 -> 13.4 | Improved |
| restrict student and matching losses to student parameters instead of updating all parameters (ablation) | KT-Net training · CRN (Car, Chair, Lamp, Sofa, Table) | CD × 10^4 (Average): 22.9 -> 13.4; 21.7 -> 13.4 | Improved |
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.
- Cycle4Completion — Cycle4Completion: Unpaired Point Cloud Completion Using Cycle Transformation With Missing Region Coding (CVPR 2021) · better on 1 of 1 shared result
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
- Cycle4Completion: Unpaired Point Cloud Completion Using Cycle Transformation With Missing Region Coding (CVPR 2021 · similarity 0.91)
- Learning Geometric Transformation for Point Cloud Completion (IJCV 2023 · similarity 0.87)
- Semantic Guided Part Relation-aware Network for Point Cloud Completion (AAAI 2026 · similarity 0.86)
- FBNet: Feedback Network for Point Cloud Completion (ECCV 2022 · similarity 0.86)
- Orthogonal Dictionary Guided Shape Completion Network for Point Cloud (AAAI 2024 · similarity 0.86)
- DinoComplete: 3D Shape Completion with Distilled Semantic Priors and State Space Models (NeurIPS 2026 · 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.
