Improving 2D Human Pose Estimation in Rare Camera Views with Synthetic Data
Not on the map. Why?
Abstract (arXiv)
Methods and datasets for human pose estimation focus predominantly on side- and front-view scenarios. We overcome the limitation by leveraging synthetic data and introduce RePoGen (RarE POses GENerator), an SMPL-based method for generating synthetic humans with comprehensive control over pose and view. Experiments on top-view datasets and a new dataset of real images with diverse poses show that adding the RePoGen data to the COCO dataset outperforms previous approaches to top- and bottom-view pose estimation without harming performance on common views. An ablation study shows that anatomical plausibility, a property prior research focused on, is not a prerequisite for effective performance. The introduced dataset and the corresponding code are available on https://mirapurkrabek.github.io/RePoGen-paper/ .
Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Builds on 27 of its 50 references that are in the map. First posted on arXiv 13 Jul 2023.
Builds on
27 of 50 references.
References · 27
- 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
- 2014Microsoft COCO: Common Objects in Contextcited by 17,608
- 2021Swin Transformer: Hierarchical Vision Transformer using Shifted Windowscited by 12,139
- 2023Adding Conditional Control to Text-to-Image Diffusion Modelscited by 5,503
- 2019Expressive Body Capture: 3D Hands, Face, and Body from a Single Imagecited by 1,612
- 2019Deep High-Resolution Representation Learning for Human Pose Estimationcited by 1,339
- 2019AMASS: Archive of Motion Capture as Surface Shapescited by 1,236
- 2022ViTPose: Simple Vision Transformer Baselines for Human Pose Estimationcited by 456
- 2017Learning from Synthetic Humanscited by 359
- 2018CrowdPose: Efficient Crowded Scenes Pose Estimation and A New Benchmarkcited by 214
- 2021AGORA: Avatars in Geography Optimized for Regression Analysiscited by 205
- 2019Distribution-Aware Coordinate Representation for Human Pose Estimationcited by 165
15 more
- 2018Pose2Seg: Detection Free Human Instance Segmentationcited by 107
- 2018Synthesizing Images of Humans in Unseen Posescited by 97
- 2022Pose-NDF: Modeling Human Pose Manifolds with Neural Distance Fieldscited by 85
- 2019The Devil is in the Details: Delving into Unbiased Data Processing for Human Pose Estimationcited by 79
- 2019Synthetic Humans for Action Recognition from Unseen Viewpointscited by 46
- 2018Jointly Optimize Data Augmentation and Network Training: Adversarial Data Augmentation in Human Pose Estimationcited by 44
- 2021Polarized Self-Attention: Towards High-quality Pixel-wise Regressioncited by 39
- 2020Adversarial Semantic Data Augmentation for Human Pose Estimationcited by 22
- 2021Let's Play for Action: Recognizing Activities of Daily Living by Learning from Life Simulation Video Gamescited by 17
- 2023Semi-Supervised 2D Human Pose Estimation Driven by Position Inconsistency Pseudo Label Correction Modulecited by 10
- 2019Unpaired Pose Guided Human Image Generationcited by 8
- 2022I^2R-Net: Intra- and Inter-Human Relation Network for Multi-Person Pose Estimationcited by 6
- 2022Humans need not label more humans: Occlusion Copy & Paste for Occluded Human Instance Segmentationcited by 4
- 2023Human Pose Estimation in Monocular Omnidirectional Top-View Imagescited by 3
- 2023Synthesizing Anyone, Anywhere, in Any Posecited by 2
Built on by
Citing papers per year: 2024: 1; 2025: 0; 2026: 1.
2 papers.