A controllable generative augmentation approach synthesizes diverse pose videos from indoor and outdoor datasets to improve model performance on unseen domains in 3D human pose estimation.
Extensive experiments show that the synthesized videos are high-quality, effective for training, and substantially improve model performance on unseen scenarios and datasets
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
Enhancing Domain Generalization in 3D Human Pose Estimation through Controllable Generative Augmentation
A controllable generative augmentation approach synthesizes diverse pose videos from indoor and outdoor datasets to improve model performance on unseen domains in 3D human pose estimation.