https://github.com/podgorskiy/stylegancpp

Unofficial implementation of StyleGAN's generator
https://github.com/podgorskiy/stylegancpp

cpp deep-learning gan generative-adversarial-network stylegan

Last synced: about 2 months ago
JSON representation

Unofficial implementation of StyleGAN's generator

• Host: GitHub
• URL: https://github.com/podgorskiy/stylegancpp
• Owner: podgorskiy
• License: apache-2.0
• Created: 2019-12-05T10:25:39.000Z (about 5 years ago)
• Default Branch: master
• Last Pushed: 2019-12-18T04:04:42.000Z (about 5 years ago)
• Last Synced: 2024-03-20T13:30:46.411Z (10 months ago)
• Topics: cpp, deep-learning, gan, generative-adversarial-network, stylegan
• Language: C++
• Homepage: http://podgorskiy.com/static/stylegan/stylegan.html
• Size: 81.1 KB
• Stars: 33
• Watchers: 6
• Forks: 7
• Open Issues: 2
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# StyleGAN C++

This is unofficial implementation of StyleGAN's generator in C++ using tensor4

* convertor.py - code that converts stylegan weights to a binary.

* compressor.cpp - code that compresses obtained binary using zfp.

* main.cpp - main part and entrypoint og the generator part.

* StyleGAN.cpp/.h - network blocks implementations.

* SConstruct - scons build script to build WebAsm version.

* CMakelists.txt - CMake script for native building.

* rgbs.pkl - retrained weights for all "to-RGB" layers except for the last one. (to produce "intermidiate" outputs)