Ferrnya / Lifting-from-the-Deep-release-for-video

Change "Lifting from the Deep: Convolutional 3D Pose Estimation from a Single Image" to video input implementation

Lifting-from-the-Deep-release-for-video

Change "Lifting from the Deep: Convolutional 3D Pose Estimation from a Single Image" to video input implementation

This project changes from Lifting from the Deep: Convolutional 3D Pose Estimation from a Single Image, CVPR 2017

Dependencies

The code is compatible with python3.5

• Tensorflow 1.0
• OpenCV

Run Video Part

• First, run setup.sh to retreive the trained models and to install the external utilities.
• Run video.py to receive the video passed from redis and save the results to the databases.
• Or run video_local.py to read the local video and save the results to the databases.

Citation

@InProceedings{Tome_2017_CVPR,
author = {Tome, Denis and Russell, Chris and Agapito, Lourdes},
title = {Lifting From the Deep: Convolutional 3D Pose Estimation From a Single Image},
booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {July},
year = {2017}
}

References

• Lifting from the Deep: Convolutional 3D Pose Estimation from a Single Image.