Temporal-evolution-inference-with-compact-feature-representation-for-talking-face-video-compression

This repository contains the source code for the paper “Beyond Keypoint Coding: Temporal Evolution Inference with Compact Feature Representation for Talking Face Video Compression” by Bolin Chen, Zhao Wang, Binzhe Li, Rongqun Lin, Shiqi Wang, and Yan Ye

The DCC keynote video can be found in https://www.youtube.com/watch?v=7en3YYT1QfU

Training

python run.py

the gpu number and save dir can be modified in run.py

Inference

python encode.py python decode.py

Reference

The training code refers to the FOMM: https://github.com/AliaksandrSiarohin/first-order-model

The arithmetic-coding refers to https://github.com/nayuki/Reference-arithmetic-coding