Rethinking on Multi-Stage Networks for Human Pose Estimation
Preprint
Abstract
Existing pose estimation approaches fall into two categories: single-stage and multi-stage methods. While multi-stage methods are seemingly more suited for the task, their performance in current practice is not as good as single-stage methods. This work studies this issue. We argue that the current multi-stage methods' unsatisfactory performance comes from the insufficiency in various design choices. We propose several improvements, including the single-stage module design, cross stage feature aggregation, and coarse-to-fine supervision. The resulting method establishes the new state-of-the-art on both MS COCO and MPII Human Pose dataset, justifying the effectiveness of a multi-stage architecture. The source code is publicly available for further research.Keywords
All Related Versions
This publication has 12 references indexed in Scilit:
- Human pose regression by combining indirect part detection and contextual informationComputers & Graphics, 2019
- OpenPose: Realtime Multi-Person 2D Pose Estimation Using Part Affinity FieldsIEEE Transactions on Pattern Analysis and Machine Intelligence, 2019
- Deep High-Resolution Representation Learning for Human Pose Estimation2019
- Multi-Scale Structure-Aware Network for Human Pose Estimation2018
- Simple Baselines for Human Pose Estimation and Tracking2018
- Cascaded Pyramid Network for Multi-person Pose Estimation2018
- Knowledge-Guided Deep Fractal Neural Networks for Human Pose EstimationIEEE Transactions on Multimedia, 2017
- Towards Accurate Multi-person Pose Estimation in the Wild2017
- Multi-context Attention for Human Pose Estimation2017
- Xception: Deep Learning with Depthwise Separable Convolutions2017