Research map / Label-efficient and robust learning / Domain adaptation and generalization
Deep Head Pose Estimation Using Synthetic Images and Partial Adversarial Domain Adaption for Continuous Label Spaces
Felix Kuhnke, Jorn Ostermann
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 |
|---|---|---|---|
| apply PADACO partial domain adaptation | BaselinePDA ResNet18 · Biwi+ (source SynHead++, partial labels) | MAE (degrees): 4.53 -> 4.13 | Improved |
| apply DANN under non-identical label spaces | BaselinePDA ResNet18 · Biwi+ (source SynHead++, partial labels) | MAE (degrees): 4.53 -> 6.05 | Hurt |
| apply PADA-like sample weighting | BaselinePDA ResNet18 · Biwi+ (source SynHead++, partial labels) | MAE (degrees): 4.53 -> 6.41 | Hurt |
| replace PADA-like distance weights with PADACO nearest-neighbor balanced resampling (ablation) | PDA with weighted random sampler · Biwi+ (source SynHead++, partial labels) | MAE (degrees): 6.41 -> 4.13 | Improved |
| replace uniform weights (DANN) with PADACO balanced resampling (ablation) | PDA ResNet18 · Biwi+ (source SynHead++, partial labels) | MAE (degrees): 6.05 -> 4.13 | Improved |
| apply DANN under shared label space | BaselineDA ResNet18 · Biwi+ (source SynBiwi+, shared labels) | MAE (degrees): 4.58 -> 3.34 | Improved |
| apply PADACO under shared label space | BaselineDA ResNet18 · Biwi+ (source SynBiwi+, shared labels) | MAE (degrees): 4.58 -> 4.04 | Improved |
| use PADACO instead of DANN when labels are shared (ablation) | DA ResNet18 · Biwi+ (source SynBiwi+, shared labels) | MAE (degrees): 3.34 -> 4.04 | Hurt |
| add domain-wise batch norm via separate source and target mini-batches | domain adaptation ResNet18 · Biwi+ | MAE (degrees) | Improved |
| add extra distance thresholding of neighbors to PADACO (ablation) | PADACO · Biwi+ | MAE (degrees) | No clear effect |
| update source weights iteratively during adversarial training (ablation) | PADACO · Biwi+ | MAE (degrees) | Hurt |
| use domain discriminator output to produce source weights (ablation) | PDA weighting · Biwi+ | MAE (degrees) | Hurt |
Closest papers by meaning
- A Dual-Augmentor Framework for Domain Generalization in 3D Human Pose Estimation (CVPR 2024 · similarity 0.88)
- Synthetic-to-Real Pose Estimation with Geometric Reconstruction (NeurIPS 2023 · similarity 0.87)
- Source-Free and Image-Only Unsupervised Domain Adaptation for Category Level Object Pose Estimation (ICLR 2024 · similarity 0.87)
- 3DGazeNet: Generalizing Gaze Estimation with Weak Supervision from Synthetic Views (ECCV 2024 · similarity 0.87)
- Prior-guided Source-free Domain Adaptation for Human Pose Estimation (ICCV 2023 · similarity 0.87)
- Lifelong Domain Adaptive 3D Human Pose Estimation (AAAI 2026 · 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.
