DragGAN-Based Emotion Image Generation and Analysis for Animated Faces

Daqi Hu

2023

Abstract

In recent years, generative artificial intelligence (AI) and its applications have become a hot topic among art designers and content creators. There is a need for a simpler and more direct method to slightly edit images. In this paper, author introduces Drag Your Generative Adversarial Network (DragGAN) and improve its discriminators and features to adapt to anime styles. This work consists of two main parts: algorithm design based on Style-Based GAN (StyleGAN) model and application to anime style images. Specifically, an analytical model is first constructed using DragGAN. The process is called motion supervision. The input image should match the trained model. Secondly, it uses point tracking to continuously iterate the generation process and gives the result of each iteration. Third, the analysis compares the predictive performance of different models and provides an interactive GUI application as a demo project. With this research, anyone can edit their anime style portraits with a few clicks and drags. This can help people to reduce the time spent on editing anime style images and increase their productivity and creativity.

Paper Citation

in Harvard Style

Hu D. (2023). DragGAN-Based Emotion Image Generation and Analysis for Animated Faces. In Proceedings of the 1st International Conference on Data Analysis and Machine Learning - Volume 1: DAML; ISBN 978-989-758-705-4, SciTePress, pages 234-239. DOI: 10.5220/0012799500003885

in Bibtex Style

@conference{daml23,
author={Daqi Hu},
title={DragGAN-Based Emotion Image Generation and Analysis for Animated Faces},
booktitle={Proceedings of the 1st International Conference on Data Analysis and Machine Learning - Volume 1: DAML},
year={2023},
pages={234-239},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0012799500003885},
isbn={978-989-758-705-4},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on Data Analysis and Machine Learning - Volume 1: DAML
TI - DragGAN-Based Emotion Image Generation and Analysis for Animated Faces
SN - 978-989-758-705-4
AU - Hu D.
PY - 2023
SP - 234
EP - 239
DO - 10.5220/0012799500003885
PB - SciTePress