Research map / Point clouds / Point cloud classification
DAP-MAE: Domain-Adaptive Point Cloud Masked Autoencoder for Effective Cross-Domain Learning
Ziqi Gao, Qiufu Li, Linlin Shen
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 cross-domain pretraining data by simple combination (ablation) | ReCon-SMC · ScanObjectNN OBJ_BG | accuracy: 94.15 -> 94.32 | No clear effect |
| add cross-domain data by simple combination | ReCon-SMC · BU3DFE expression recognition | accuracy: 89.13 -> 88.52 | Hurt |
| add cross-domain data by simple combination | ReCon-SMC · Bosphorus expression recognition | accuracy: 87.69 -> 87.23 | Hurt |
| add cross-domain data by simple combination | ReCon-SMC · ScanNetV2 detection | AP50: 42.7 -> 42.5 | Hurt |
| add heterogeneous domain adapter (ablation) | DAP-MAE with cross-domain data · ScanObjectNN OBJ_BG | accuracy: 94.32 -> 94.66 | Improved |
| add domain feature generator (ablation) | DAP-MAE with cross-domain data · ScanObjectNN OBJ_BG | accuracy: 94.32 -> 94.66 | Improved |
| add both HDA and DFG (ablation) | DAP-MAE with cross-domain data · ScanObjectNN OBJ_BG | accuracy: 94.32 -> 95.18 | Improved |
| add fusion of other-domain MLPs in HDA (ablation) | DAP-MAE · ScanObjectNN OBJ_BG | accuracy: 94.66 -> 95.18 | Improved |
| use MLP instead of plain adding to fuse domains O+F (ablation) | DAP-MAE · ScanObjectNN OBJ_BG | accuracy: 92.59 -> 93.80 | Improved |
| use MLP instead of FC to predict fusion coefficients for O+F+S (ablation) | DAP-MAE · ScanObjectNN OBJ_BG | accuracy: 93.80 -> 95.18 | Improved |
| combine class, domain and point cloud features in head (ablation) | DAP-MAE · ScanObjectNN OBJ_BG | accuracy: 94.49 -> 95.18 | Improved |
| use loss weights 100/0.001 instead of 1.0/1.0 (ablation) | DAP-MAE · ScanObjectNN OBJ_BG | accuracy: 93.80 -> 95.18 | Improved |
Closest papers by meaning
- 3D Feature Prediction for Masked-AutoEncoder-Based Point Cloud Pretraining (ICLR 2024 · similarity 0.89)
- PCP-MAE: Learning to Predict Centers for Point Masked Autoencoders (NeurIPS 2024 · similarity 0.88)
- Towards Compact 3D Representations via Point Feature Enhancement Masked Autoencoders (AAAI 2024 · similarity 0.87)
- RI-MAE: Rotation-Invariant Masked AutoEncoders for Self-Supervised Point Cloud Representation Learning (AAAI 2025 · similarity 0.86)
- DAPoinTr: Domain Adaptive Point Transformer for Point Cloud Completion (AAAI 2025 · similarity 0.86)
- PointDGMamba: Domain Generalization of Point Cloud Classification via Generalized State Space Model (AAAI 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.
