Research map / 3D vision / Human pose estimation
3D Human Pose Estimation in the Wild by Adversarial Learning
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 adversarial learning with the full multi-source discriminator (ablation) | Human3.6M pose estimator · Human3.6M Protocol #1 | MPJPE (mm): 64.8 -> 59.7 | Helped |
| add pairwise geometric descriptors to image input for adversarial learning (ablation) | Human3.6M pose estimator · Human3.6M Protocol #1 | MPJPE (mm): 64.8 -> 60.3 | Helped |
| use heatmaps and depth maps rather than geometric descriptors alongside the image (ablation) | Human3.6M pose estimator · Human3.6M Protocol #1 | MPJPE (mm): 64.8 -> 61.3 | Helped |
| add geometric descriptors to the heatmap/depth-map discriminator sources (ablation) | Human3.6M pose estimator · Human3.6M Protocol #1 | MPJPE (mm): 61.3 -> 59.7 | Helped |
| pretrain the depth regressor before adversarial learning (ablation) | Full adversarial pose estimator · Human3.6M Protocol #1 | MPJPE (mm): 63.4 -> 59.7 | Helped |
| fine-tune the 2D pose module end-to-end instead of fixing it (ablation) | Full adversarial pose estimator · Human3.6M Protocol #1 | MPJPE (mm): 63.1 -> 59.7 | Helped |
| fine-tune the 2D pose module instead of fixing it (ablation) | Baseline pose estimator · Human3.6M Protocol #1 | MPJPE (mm): 65.2 -> 64.8 | No clear effect |
| apply adversarial learning to the pretrained 2D pose module (ablation) | MPII 2D pose estimator · MPII validation | PCKh@0.5: 87.6 -> 88.6 | Helped |
| apply adversarial learning using Human3.6M and MPII training data | 3D pose estimator · MPI-INF-3DHP test | PCK: 50.1 -> 69.0 | Helped |
| apply adversarial learning using Human3.6M and MPII training data | 3D pose estimator · MPI-INF-3DHP test | AUC: 21.6 -> 32.0 | Helped |
Closest papers by meaning
- In the Wild Human Pose Estimation Using Explicit 2D Features and Intermediate 3D Representations (CVPR 2019 · similarity 0.93)
- 3D Human Pose Estimation From a Single Image via Distance Matrix Regression (CVPR 2017 · similarity 0.91)
- Coarse-To-Fine Volumetric Prediction for Single-Image 3D Human Pose (CVPR 2017 · similarity 0.90)
- Kinematic-Structure-Preserved Representation for Unsupervised 3D Human Pose Estimation (AAAI 2020 · similarity 0.89)
- A Dual-Augmentor Framework for Domain Generalization in 3D Human Pose Estimation (CVPR 2024 · similarity 0.89)
- RepNet: Weakly Supervised Training of an Adversarial Reprojection Network for 3D Human Pose Estimation (CVPR 2019 · similarity 0.89)
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.