Human Pose Regression by Combining Indirect Part Detection and Contextual Information

Abstract

In this paper, we propose an end-to-end trainable regression approach for human pose estimation from still images. We use the proposed Soft-argmax function to convert feature maps directly to joint coordinates, resulting in a fully differentiable framework. Our method is able to learn heat maps representations indirectly, without additional steps of artificial ground truth generation. Consequently, contextual information can be included to the pose predictions in a seamless way. We evaluated our method on two very challenging datasets, the Leeds Sports Poses (LSP) and the MPII Human Pose datasets, reaching the best performance among all the existing regression methods and comparable results to the state-of-the-art detection based approaches.

References in corpus (13)

• DeepPose: Human Pose Estimation via Deep Neural Networks
• Human pose estimation via Convolutional Part Heatmap Regression
• Convolutional Pose Machines
• DeeperCut: A Deeper, Stronger, and Faster Multi-Person Pose Estimation Model
• Multi-Context Attention for Human Pose Estimation
• LIFT: Learned Invariant Feature Transform
• Combining Local Appearance and Holistic View: Dual-Source Deep Neural Networks for Human Pose Estimation
• Flowing ConvNets for Human Pose Estimation in Videos
• Learning Feature Pyramids for Human Pose Estimation
• Adversarial PoseNet: A Structure-aware Convolutional Network for Human Pose Estimation
• Compositional Human Pose Regression
• Self Adversarial Training for Human Pose Estimation
• 3D Human Pose Estimation with 2D Marginal Heatmaps

Cited by in corpus (28)

• Monocular Human Pose Estimation: A Survey of Deep Learning-based Methods
• Numerical Coordinate Regression with Convolutional Neural Networks
• Multi-task Deep Learning for Real-Time 3D Human Pose Estimation and Action Recognition
• Rethinking on Multi-Stage Networks for Human Pose Estimation
• MeTRAbs: Metric-Scale Truncation-Robust Heatmaps for Absolute 3D Human Pose Estimation
• Deep Learning-Based Human Pose Estimation: A Survey
• Human Body Pose Estimation for Gait Identification: A Comprehensive Survey of Datasets and Models
• Wide-Baseline Relative Camera Pose Estimation with Directional Learning
• Heatmap Regression via Randomized Rounding
• Simple and Lightweight Human Pose Estimation
• Alpha-Refine: Boosting Tracking Performance by Precise Bounding Box Estimation
• SMPLR: Deep SMPL reverse for 3D human pose and shape recovery
• Metric-Scale Truncation-Robust Heatmaps for 3D Human Pose Estimation
• Integral Human Pose Regression
• U-BEV: Height-aware Bird's-Eye-View Segmentation and Neural Map-based Relocalization
• DeepFuse: An IMU-Aware Network for Real-Time 3D Human Pose Estimation from Multi-View Image
• 3D Human Pose Estimation with 2D Marginal Heatmaps
• DeepSkeleton: Skeleton Map for 3D Human Pose Regression
• Anti-Confusing: Region-Aware Network for Human Pose Estimation
• FlatteNet: A Simple Versatile Framework for Dense Pixelwise Prediction
• Multi-Person Pose Estimation with Enhanced Feature Aggregation and Selection
• KPNet: Towards Minimal Face Detector
• Separable Batch Normalization for Robust Facial Landmark Localization with Cross-protocol Network Training
• SDR-GAIN: A High Real-Time Occluded Pedestrian Pose Completion Method for Autonomous Driving
• CLTA: Contents and Length-based Temporal Attention for Few-shot Action Recognition
• Alleviating Human-level Shift : A Robust Domain Adaptation Method for Multi-person Pose Estimation
• Domain Adaptation of Learned Features for Visual Localization
• Low-resolution Human Pose Estimation