Cascaded Deep Monocular 3D Human Pose Estimation With Evolutionary Training Data
Authors:
- 1 June 2020
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 6172-6182
Abstract
End-to-end deep representation learning has achieved remarkable accuracy for monocular 3D human pose estimation, yet these models may fail for unseen poses with limited and fixed training data. This paper proposes a novel data augmentation method that: (1) is scalable for synthesizing massive amount of training data (over 8 million valid 3D human poses with corresponding 2D projections) for training 2D-to-3D networks, (2) can effectively reduce dataset bias. Our method evolves a limited dataset to synthesize unseen 3D human skeletons based on a hierarchical human representation and heuristics inspired by prior knowledge. Extensive experiments show that our approach not only achieves state-of-the-art accuracy on the largest public benchmark, but also generalizes significantly better to unseen and rare poses. Relevant files and tools are available at the project website.Keywords
All Related Versions
This publication has 56 references indexed in Scilit:
- SMPLACM Transactions on Graphics, 2015
- Pose-conditioned joint angle limits for 3D human pose reconstruction2015
- 3D Pictorial Structures for Multiple Human Pose Estimation2014
- Face Alignment at 3000 FPS via Regressing Local Binary Features2014
- Face Alignment by Explicit Shape RegressionInternational Journal of Computer Vision, 2013
- Human3.6M: Large Scale Datasets and Predictive Methods for 3D Human Sensing in Natural EnvironmentsIEEE Transactions on Pattern Analysis and Machine Intelligence, 2013
- 3D Pictorial Structures for Multiple View Articulated Pose Estimation2013
- Unbiased look at dataset bias2011
- HumanEva: Synchronized Video and Motion Capture Dataset and Baseline Algorithm for Evaluation of Articulated Human MotionInternational Journal of Computer Vision, 2009
- SCAPE2005