Replica is a motion-transfer based project. We read about the idea first from the paper "Everybody Dance Now" We are following their general approach, with modyifing the algorithms of each module to which suits us best.

It contains on multiple main modules:

• Image Segmentation
• Pose Estimation
• Generative Addverersial Networks (GAN)

Everybody Dance Now dataset is used for training our GAN model in developing phase. After this, we uses our own dataset generated from input videos for training.

• Radwa Samy Khattab
• Salma Ibrahim Ismail
• Mariam Hesham Heiba
• Nehal AbdelKader Bayoumi

• Everybody Dance Now paper
• Human Pose Estimation and Tracking via Parsing a Tree Structure Based Human Model

This project is our graduation project for Computer Engineering degree from Cairo University, in 2021.