Need help with selfie2anime?
Click the “chat” button below for chat support from the developer who created it, or find similar developers for support.

About the developer

t04glovern
153 Stars 23 Forks MIT License 43 Commits 0 Opened issues

Description

Anime2Selfie Backend Services - Lambda, Queue, API Gateway and traffic processing

Services available

!
?

Need anything else?

Contributors list

# 88,668
python-...
Tensorf...
tf2
yolov3
41 commits
# 374,810
data-sc...
aws-lam...
Serverl...
HTML
1 commit

Selfie2Anime

Selfie2Anime  Version  Status

Selfie2Anime

What do YOU look like in ANIME?

This repository contains the source code for the backend for the selfie2anime.com website.

Source code for the frontend website can be found at https://github.com/SilentByte/selfie2anime-site

Also checkout the presentation "Scaling Models to the Masses" I did for Perth Machine Learning Group


How Does it Work?


Using machine learning techniques combined with a Generative Adversarial Network (GAN) makes it possible to generate anime-style characters based on real people. With this website, you can generate your own anime alter ego!

The GAN we are using is based on original work by Junho Kim, Minjae Kim, Hyeonwoo Kang, and Kwanghee Lee. More information can be found in their awesome repository, which is available here, or in their research paper.


Components


Below is a general diagram illustrating the process our workers follow to process incoming requests.

More information about the process seen above can be found in the following modules

Architecture Diagram

  • image-handler
    • Image processing pipeline
  • UGATIT
    • UGATIT environment from Junho Kim and Minjae Kim and Hyeonwoo Kang and Kwanghee Lee

Attributions

Please cite the original author of UGATIT:

  • Junho Kim, Minjae Kim, Hyeonwoo Kang, Kwanghee Lee
@misc{kim2019ugatit,
    title={U-GAT-IT: Unsupervised Generative Attentional Networks with Adaptive Layer-Instance Normalization for Image-to-Image Translation},
    author={Junho Kim and Minjae Kim and Hyeonwoo Kang and Kwanghee Lee},
    year={2019},
    eprint={1907.10830},
    archivePrefix={arXiv},
    primaryClass={cs.CV}
}

We use cookies. If you continue to browse the site, you agree to the use of cookies. For more information on our use of cookies please see our Privacy Policy.