Research map / 3D vision / Human pose estimation
MemPose: Category-level Object Pose Estimation with Memory
Xiao Lin, Minghao Zhu, Yun Peng, Liuyi Wang, Qiyi Wang, Chengju Liu, Qijun Chen
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 category-level memory module to AG-Pose-style baseline (ablation) | MemPose · REAL275 | 5°2cm: 57.0 -> 59.9 | Improved |
| add memory module (ablation) | MemPose · REAL275 | 10°2cm: 75.1 -> 79.0 | Improved |
| increase buffer length from 96 to 256 (ablation) | MemPose · REAL275 | 5°2cm: 59.9 -> 59.1 | No clear effect |
| store pooled local features instead of global features (ablation) | MemPose · REAL275 | 5°2cm: 58.0 -> 59.9 | Improved |
| buffer warm-up before update (ablation) | MemPose · REAL275 | 5°5cm: 66.0 -> 67.7 | Improved |
| update memory before retrieval instead of after (ablation) | MemPose · REAL275 | 5°5cm: 65.2 -> 67.7 | Improved |
| similarity-based merge with Gumbel noise instead of FIFO (ablation) | MemPose · REAL275 | 5°2cm: 59.1 -> 59.9 | Improved |
| Gumbel noise balancing instead of top-k balancing (ablation) | MemPose · REAL275 | 5°5cm: 65.3 -> 67.7 | Improved |
| retrieve all memory entries instead of top-16 (ablation) | MemPose · REAL275 | 5°2cm: 57.9 -> 59.9 | Improved |
| gated fusion instead of addition or concatenation (ablation) | MemPose · REAL275 | 5°5cm: 65.3 -> 67.7 | Improved |
| replace memory module with parameter-matched MLPs (ablation) | MemPose · REAL275 | 5°2cm: 59.9 -> 57.8 | Hurt |
| update memory during inference (test stream) | MemPose · REAL275 | 5°2cm: 59.9 -> 63.6 | Improved |
Closest papers by meaning
- PriorPose: Reference-Guided Joint Deformation and Alignment for Category-Level Object Pose Estimation (ECCV 2026 · similarity 0.91)
- Pose Prior Learner: Unsupervised Categorical Prior Learning for Pose Estimation (ICLR 2026 · similarity 0.89)
- Unsupervised Learning of Category-Level 3D Pose from Object-Centric Videos (CVPR 2024 · similarity 0.89)
- SchemaPose: RGB-Based Category-Level Pose Estimation with Parametric Category Schema (NeurIPS 2026 · similarity 0.88)
- SecondPose: SE(3)-Consistent Dual-Stream Feature Fusion for Category-Level Pose Estimation (CVPR 2024 · similarity 0.88)
- GeoReF: Geometric Alignment Across Shape Variation for Category-level Object Pose Refinement (CVPR 2024 · similarity 0.87)
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.
