MoDeep: A Deep Learning Framework Using Motion Features for Human Pose Estimation

Abstract

In this work, we propose a novel and efficient method for articulated human pose estimation in videos using a convolutional network architecture, which incorporates both color and motion features. We propose a new human body pose dataset, FLIC-motion, that extends the FLIC dataset with additional motion features. We apply our architecture to this dataset and report significantly better performance than current state-of-the-art pose detection systems.

References in corpus (4)

• DeepPose: Human Pose Estimation via Deep Neural Networks
• Visualizing and Understanding Convolutional Networks
• Learning Human Pose Estimation Features with Convolutional Networks
• Fast Image Scanning with Deep Max-Pooling Convolutional Neural Networks

Cited by in corpus (14)

• Stacked Hourglass Networks for Human Pose Estimation
• Deep learning tools for the measurement of animal behavior in neuroscience
• A Primer on Motion Capture with Deep Learning: Principles, Pitfalls and Perspectives
• Learning Spatiotemporal Features with 3D Convolutional Networks
• Flowing ConvNets for Human Pose Estimation in Videos
• Recurrent Network Models for Human Dynamics
• Human Pose and Path Estimation from Aerial Video using Dynamic Classifier Selection
• Monocular 3D Human Pose Estimation In The Wild Using Improved CNN Supervision
• Thin-Slicing Network: A Deep Structured Model for Pose Estimation in Videos
• PoseTrack: Joint Multi-Person Pose Estimation and Tracking
• EgoCap: Egocentric Marker-less Motion Capture with Two Fisheye Cameras
• Real-time Human Pose Estimation from Video with Convolutional Neural Networks
• Bi-directional Graph Structure Information Model for Multi-Person Pose Estimation
• Keep it SMPL: Automatic Estimation of 3D Human Pose and Shape from a Single Image