Learning from Synthetic Humans

01/05/2017
by   Gul Varol, et al.
0

Estimating human pose, shape, and motion from images and videos are fundamental challenges with many applications. Recent advances in 2D human pose estimation use large amounts of manually-labeled training data for learning convolutional neural networks (CNNs). Such data is time consuming to acquire and difficult to extend. Moreover, manual labeling of 3D pose, depth and motion is impractical. In this work we present SURREAL (Synthetic hUmans foR REAL tasks): a new large-scale dataset with synthetically-generated but realistic images of people rendered from 3D sequences of human motion capture data. We generate more than 6 million frames together with ground truth pose, depth maps, and segmentation masks. We show that CNNs trained on our synthetic dataset allow for accurate human depth estimation and human part segmentation in real RGB images. Our results and the new dataset open up new possibilities for advancing person analysis using cheap and large-scale synthetic data.

READ FULL TEXT

page 1

page 2

page 3

page 5

page 7

page 8

research
04/17/2023

Human Pose Estimation in Monocular Omnidirectional Top-View Images

Human pose estimation (HPE) with convolutional neural networks (CNNs) fo...
research
05/25/2016

Real-Time Human Motion Capture with Multiple Depth Cameras

Commonly used human motion capture systems require intrusive attachment ...
research
06/29/2023

BEDLAM: A Synthetic Dataset of Bodies Exhibiting Detailed Lifelike Animated Motion

We show, for the first time, that neural networks trained only on synthe...
research
12/23/2021

HSPACE: Synthetic Parametric Humans Animated in Complex Environments

Advances in the state of the art for 3d human sensing are currently limi...
research
09/30/2016

A deep representation for depth images from synthetic data

Convolutional Neural Networks (CNNs) trained on large scale RGB database...
research
07/04/2019

Sim2real transfer learning for 3D pose estimation: motion to the rescue

Simulation is an anonymous, low-bias source of data where annotation can...
research
10/30/2019

Real-time Convolutional Networks for Depth-based Human Pose Estimation

We propose to combine recent Convolutional Neural Networks (CNN) models ...

Please sign up or login with your details

Forgot password? Click here to reset