Research map / 3D vision / Human pose estimation
HopFIR: Hop-wise GraphFormer with Intragroup Joint Refinement for 3D Human Pose Estimation
Kai Zhai, Qiang Nie, Bo Ouyang, Xiang Li, Shanlin 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 |
|---|---|---|---|
| add HGF module to HopGCN (ablation) | HopGCN · Human3.6M (GT 2D) | MPJPE (mm): 39.15 -> 35.19 | Improved |
| add IJR module to HopGCN (ablation) | HopGCN · Human3.6M (GT 2D) | MPJPE (mm): 39.15 -> 36.62 | Improved |
| combine HGF and IJR (HopFIR block) (ablation) | HopGCN · Human3.6M (GT 2D) | MPJPE (mm): 35.19 -> 32.67 | Improved |
| use hop-wise attention instead of plain transformer encoder over joints (ablation) | HopFIR · Human3.6M (GT 2D) | MPJPE (mm): 36.21 -> 32.67 | Improved |
| use skeleton graph prior instead of random graph (ablation) | HopFIR · Human3.6M (GT 2D) | MPJPE (mm): 34.68 -> 32.67 | Improved |
| insert HA layer into SemGCN | SemGCN · Human3.6M (GT 2D) | MPJPE (mm): 42.14 -> 38.41 | Improved |
| insert HA layer into Modulated GCN | Modulated GCN · Human3.6M (GT 2D) | MPJPE (mm): 38.25 -> 36.54 | Improved |
| increase k hops from 1 to 3 (ablation) | HopFIR · Human3.6M (GT 2D) | MPJPE (mm): 35.88 -> 32.67 | Improved |
| increase k hops from 3 to 4 (ablation) | HopFIR · Human3.6M (GT 2D) | MPJPE (mm): 32.67 -> 35.58 | Hurt |
| increase blocks from 3 to 4 (ablation) | HopFIR · Human3.6M (GT 2D) | MPJPE (mm): 32.67 -> 33.84 | Hurt |
| place IJR last, (H)(H)(I), instead of first (ablation) | HopFIR · Human3.6M (GT 2D) | MPJPE (mm): 34.20 -> 32.67 | Improved |
| reduce channels to 64 (ablation) | HopFIR · Human3.6M (GT 2D) | MPJPE (mm): 32.67 -> 33.52 | Mixed |
Later papers that compared with this method
1 later accepted paper reports HopFIR as a baseline in their results tables; 0 of them report a better number on most of the results they share with it.
- Sequential Joint Dependency Aware Human Pose Estimation with State Space Model (AAAI 2025 · better on 2 of 4 shared results)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 on most shared results.
- MixSTE — MixSTE: Seq2seq Mixed Spatio-Temporal Encoder for 3D Human Pose Estimation in Video (CVPR 2022) · better on 0 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
- Modulated Graph Convolutional Network for 3D Human Pose Estimation (ICCV 2021 · similarity 0.88)
- Pose-Oriented Transformer with Uncertainty-Guided Refinement for 2D-to-3D Human Pose Estimation (AAAI 2023 · similarity 0.88)
- Learning Skeletal Graph Neural Networks for Hard 3D Pose Estimation (ICCV 2021 · similarity 0.88)
- Deep Semantic Graph Transformer for Multi-View 3D Human Pose Estimation (AAAI 2024 · similarity 0.88)
- HiPoser: 3D Human Pose Estimation with Hierarchical Shared Learning at Parts-Level Using Inertial Measurement Units (AAAI 2025 · similarity 0.86)
- SVTformer: Spatial-View-Temporal Transformer for Multi-View 3D Human Pose Estimation (AAAI 2025 · 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.
