Research map / 3D vision / Human pose estimation
HccePose(BF): Predicting Front & Back Surfaces to Construct Ultra-Dense 2D-3D Correspondences for Pose Estimation
Yulin Wang, Mengting Hu, Hongli Li, Chen Luo
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 hierarchical binary coordinate encoding (HBCE) in place of ZebraPose surface-region encoding (ablation) | ZebraPose · IC-BIN | AR of ADD(-S): 55.85 -> 56.82 | Improved |
| replace HBCE with hierarchical continuous coordinate encoding (HCCE, h3) (ablation) | HBCE · IC-BIN | AR of ADD(-S): 56.82 -> 61.95 | Improved |
| add hierarchy to continuous coordinate encoding (CCE to HCCE) (ablation) | CCE · IC-BIN | AR of ADD(-S): 55.42 -> 61.35 | Improved |
| add single-histogram loss weight adjustment (h1) to HCCE (ablation) | HCCE · IC-BIN | AR of ADD(-S): 61.35 -> 60.44 | Hurt |
| add multi-histogram loss weight adjustment (h3) to HCCE (ablation) | HCCE · IC-BIN | AR of ADD(-S): 61.35 -> 61.95 | Improved |
| add multi-histogram weight adjustment (h3) over single histogram (h1) (ablation) | HCCE · IC-BIN | AR of ADD-S: 65.23 -> 68.30 | Improved |
| predict back surface in addition to front surface (bf) (ablation) | front-only (f) · LM-O/TUD-L/IC-BIN | AP of BOP score: 82.2 -> 82.6 | Improved |
| predict front surface in addition to back surface (bf) (ablation) | back-only (b) · LM-O/TUD-L/IC-BIN | AP of BOP score: 81.3 -> 82.6 | Improved |
| add uniform sampling between front and back surfaces (bfu) (ablation) | bf · LM-O/TUD-L/IC-BIN | AP of BOP score: 82.6 -> 83.3 | Improved |
| use ultra-dense correspondences (bfu) instead of front only (ablation) | front-only (f) · LM-O/TUD-L/IC-BIN | AP of BOP score: 82.2 -> 83.3 | Improved |
| HccePose(BF) vs ZebraPose on RGB 6D localization | ZebraPose · BOP core (RGB) | AR(AP) Core: 74.9 -> 77.3 | Improved |
| HccePose(BF) vs ZebraPose on 2D segmentation | ZebraPose · BOP core (RGB) | AR(AP) Core: 61.9 -> 65.6 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- ZebraPose — ZebraPose: Coarse To Fine Surface Encoding for 6DoF Object Pose Estimation (CVPR 2022) · better on 8 of 9 shared results
- GDRNPP — GDRNPP: A Geometry-Guided and Fully Learning-Based Object Pose Estimator (TPAMI 2025) · 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
- UltraPose: Synthesizing Dense Pose With 1 Billion Points by Human-Body Decoupling 3D Model (ICCV 2021 · similarity 0.86)
- SimPose: Effectively Learning DensePose and Surface Normals of People from Simulated Data (ECCV 2020 · similarity 0.86)
- Learning 3D Human Shape and Pose from Dense Body Parts (TPAMI 2020 · similarity 0.86)
- Front2Back: Single View 3D Shape Reconstruction via Front to Back Prediction (CVPR 2020 · similarity 0.85)
- HiPose: Hierarchical Binary Surface Encoding and Correspondence Pruning for RGB-D 6DoF Object Pose Estimation (CVPR 2024 · similarity 0.84)
- HiPART: Hierarchical Pose AutoRegressive Transformer for Occluded 3D Human Pose Estimation (CVPR 2025 · similarity 0.84)
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.
