PeterWang512/CNNDetection
A deep learning classifier for detecting images generated by CNNs such as StyleGAN.
Not currently ranked — collecting fresh signals.
star history
This repository contains the models, evaluation code, and training code from a CVPR 2020 paper on detecting CNN-generated images. It provides a pretrained binary classifier that distinguishes real photographs from synthetic images generated by GANs. The project includes tools for evaluating detection performance on various GAN architectures including StyleGAN2 and StyleGAN3.
Frequently asked
- What is PeterWang512/CNNDetection?
- A deep learning classifier for detecting images generated by CNNs such as StyleGAN.
- Is CNNDetection open source?
- Yes — PeterWang512/CNNDetection is an open-source project tracked on heatdrop.
- What language is CNNDetection written in?
- PeterWang512/CNNDetection is primarily written in Python.
- How popular is CNNDetection?
- PeterWang512/CNNDetection has 1k stars on GitHub.
- Where can I find CNNDetection?
- PeterWang512/CNNDetection is on GitHub at https://github.com/PeterWang512/CNNDetection.