StyleGAN

## Overview

StyleGAN is a dataset for classification tasks - it contains Generate Images annotations for 1,881 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [CC BY 4.0 license](https://creativecommons.org/licenses/CC BY 4.0).

Dataset

This dataset was created by Cường Phạm 281205

Contents

Comparison of InceptionResnetV2-StyleGAN2-Augmented and InceptionResNetV2-Conventional-Augmentation performance evaluation.

A dataset for analyzing and improving the image quality of StyleGAN.

Dataset

This dataset was created by anant jain git

Contents

Dataset

This dataset was created by Tommy NgX

Contents

This dataset of natural images, as utilized in this notebook for neural decoding, forms part of the Brain2GAN study. It comprises w-latents of visual stimuli and corresponding (normalized) multi-unit activity (MUA) responses from the macaque visual cortex. The w-latents should be fed to the generator of StyleGAN-XL to recreate the visual stimuli. Note that the w-latents of the training and test set are included in separate .npy files.

TrueFace is a first dataset of social media processed real and synthetic faces, obtained by the successful StyleGAN generative models, and shared on Facebook, Twitter and Telegram.

Images have historically been a universal and cross-cultural communication medium, capable of reaching people of any social background, status or education. Unsurprisingly though, their social impact has often been exploited for malicious purposes, like spreading misinformation and manipulating public opinion. With today's technologies, the possibility to generate highly realistic fakes is within everyone's reach. A major threat derives in particular from the use of synthetically generated faces, which are able to deceive even the most experienced observer. To contrast this fake news phenomenon, researchers have employed artificial intelligence to detect synthetic images by analysing patterns and artifacts introduced by the generative models. However, most online images are subject to repeated sharing operations by social media platforms. Said platforms process uploaded images by applying operations (like compression) that progressively degrade those useful forensic traces, compromising the effectiveness of the developed detectors. To solve the synthetic-vs-real problem "in the wild", more realistic image databases, like TrueFace, are needed to train specialised detectors.

See https://github.com/neuralcodinglab/brain2gan.

Introduction

Recent advances in generative adversarial networks (GANs) have led to remarkable achievements in face image synthesis. While methods that use style-based GANs can generate strikingly photorealistic face images, it is often difficult to control the characteristics of the generated faces in a meaningful and disentangled way. Prior approaches aim to achieve such semantic control and disentanglement within the latent space of a previously trained GAN. In contrast, we propose a framework that a priori models physical attributes of the face such as 3D shape, albedo, pose, and lighting explicitly, thus providing disentanglement by design. Our method, MOST-GAN, integrates the expressive power and photorealism of style-based GANs with the physical disentanglement and flexibility of nonlinear 3D morphable models, which we couple with a state-of-the-art 2D hair manipulation network. MOST-GAN achieves photorealistic manipulation of portrait images with fully disentangled 3D control over their physical attributes, enabling extreme manipulation of lighting, facial expression, and pose variations up to full profile view.

To foster further research into this topic, we are publicly releasing our pre-trained model for MOST-GAN. Please see our AAAI paper titled [MOST-GAN: 3D Morphable StyleGAN for Disentangled Face Image Manipulation](https://arxiv.org/abs/2111.01048) for details.

At a Glance

-The size of the unzipped model is ~300MB.

-The unzipped folder contains: (i) a README.md file and (ii) ./checkpoints/checkpoint01.pt pre-trained model. The pre-trained model could be loaded in our publicly released MOST-GAN implementation.

Citation

If you use the MOST-GAN data in your research, please cite our paper:

@inproceedings{medin2022most,
 title={MOST-GAN: 3D morphable StyleGAN for disentangled face image manipulation},
 author={Medin, Safa C and Egger, Bernhard and Cherian, Anoop and Wang, Ye and Tenenbaum, Joshua B and Liu, Xiaoming and Marks, Tim K},
 booktitle={Proceedings of the AAAI conference on artificial intelligence},
 volume={36},
 number={2},
 pages={1962--1971},
 year={2022}
}

License

The MOST-GAN data is released under CC-BY-SA-4.0 license.

All data:

Created by Mitsubishi Electric Research Laboratories (MERL), 2022,2023
 
SPDX-License-Identifier: CC-BY-SA-4.0

Real vs Fake Image Dataset

Overview

This dataset consists of two primary categories: real_images and fake_images. The real_images category contains authentic images, while the fake_images category includes synthetic images generated using various advanced generative models. The purpose of this dataset is to facilitate research and development in the field of image classification, focusing on distinguishing between real and synthetic images.

Dataset Structure

The dataset is organized as follows:

fake_images

The fake_images folder contains synthetic images generated using various generative models. Each subfolder represents a specific image generation model:

• big_gan: Images generated using the BigGAN model.
• cips: Images generated by CIPS (Conditional Image Prior Sampling).
• ddpm: Images generated by Denoising Diffusion Probabilistic Models.
• denoising_diffusion_gan: Hybrid GAN and diffusion model.
• diffusion_gan: GANs using diffusion processes for image generation.
• face_synthetics: Synthetic face images generated using models like StyleGAN.
• gansformer: GAN-based transformer architecture for image synthesis.
• gau_gan: Images generated from sketches.
• generative_inpainting: Images generated via inpainting.
• glide: Text-to-image generative model.
• lama: Latent manifold-based image generation.
• latent_diffusion: Diffusion model operating in latent space.
• mat: Artistic texture generation model.
• palette: Colorful image generation model.
• projected_gan: GANs with projected approaches for quality improvements.
• sfhq: High-resolution synthetic facial images.
• stable_diffusion: Popular image generation using stable diffusion models.
• star_gan: Multi-domain image transformation.
• stylegan1: First version of the StyleGAN architecture.
• stylegan2: Improved version of StyleGAN.
• stylegan3: Latest version of StyleGAN with more stable and realistic output.
• taming_transformer: Transformer-based image generation.
• vq_diffusion: Model combining vector quantization with diffusion.

real_images

This folder contains authentic, real-world images, which are used as the ground truth for comparison with the generated fake_images.

Usage

This dataset can be used for training and evaluating image classification models, particularly those focused on distinguishing real images from synthetic ones. It is well-suited for experiments with generative adversarial networks (GANs), diffusion models, and other deep learning techniques.

Augmented images from Images data augmentation for industry applications through StyleGAN and transfer learning study

Flickr-Faces-HQ (FFHQ) is a high-quality image dataset of human faces, originally created as a benchmark for generative adversarial networks (GAN). The dataset consists of 70,000 high-quality PNG images at 1024x1024 resolution and contains considerable variation in terms of age, ethnicity and image background. It also has good coverage of accessories such as eyeglasses, sunglasses, hats, etc. The images were crawled from Flickr, thus inheriting all the biases of that website, and automatically aligned and cropped using dlib. Only images under permissive licenses were collected. Various automatic filters were used to prune the set, and finally Amazon Mechanical Turk was used to remove the occasional statues, paintings, or photos of photos.

This dataset was created to test whether it's possible to build a general-purpose detector that can tell real images apart from fake ones generated by convolutional neural networks (CNNs), no matter which model or dataset was used to create the fake images.

To do this, the authors collected fake images generated by 11 different CNN-based image generation models. These models represent a wide range of current image synthesis techniques and include:

ProGAN

StyleGAN

BigGAN

CycleGAN

StarGAN

GauGAN

DeepFakes

Cascaded Refinement Networks (CRN)

Implicit Maximum Likelihood Estimation (IMLE)

Second-order Attention Super-Resolution (SOAT-SR)

Seeing-in-the-Dark (SID)

The dataset includes fake images from each of these models and a set of real images, allowing for binary classification (real vs. fake).

The study found that a standard image classifier (like a convolutional neural network) trained on fake images from just one generator (ProGAN) was able to detect fake images from other, completely different generators with surprising accuracy. This suggests that many CNN-generated images, even from different architectures, share common flaws that can be learned and detected.

The dataset is useful for research in detecting synthetic media, improving image forensics, and understanding the weaknesses in current generative models.

Code and pre-trained models were made available by the authors (https://github.com/chuangchuangtan/GANGen-Detection).

Parameters used in training StyleGAN2 model, during training a snapshot of the model is saved every 10 ticks.

Microscopic wood identification plays a critical role in many economically important areas in wood science. Historically, producing and curating relevant and representative microscopic cross-section images of wood species is limited to highly experienced and trained anatomists. This manuscript demonstrates the feasibility of generating synthetic microscopic cross-sections of hardwood species. We leveraged a publicly available dataset of 119 hardwood species to train a style-based generative adversarial network (GAN). The proposed GAN generated anatomically accurate cross-section images with remarkable fidelity to actual data. Quantitative metrics corroborated the capacity of the generative model in capturing complex wood structure by resulting in a Fréchet inception distance score of 17.38. Image diversity was calculated using the Structural Similarity Index Measure (SSIM). The SSIM results confirmed that the GAN approach can successfully synthesize diverse images. To confirm the usefulness and realism of the GAN generated images, eight professional wood anatomists in two experience levels participated in a visual Turing test and correctly identified fake and actual images at rates of 48.3 and 43.7%, respectively, with no statistical difference when compared to random guess. The generative model can synthesize realistic, diverse, and meaningful high-resolution microscope cross-section images that are virtually indistinguishable from real images. Furthermore, the framework presented may be suitable for improving current deep learning models, helping understand potential breeding between species, and may be used as an educational tool.

1. Overview

Filtered 4,703 Good Quality Female images generated by GenForce StyleGAN celeba_partial-256x256, for Oh-LoRA Project.

Filtered by Gender and Quality CNN Models (Link)

  2. Save Path & Related Link

Save Path 2025_04_08_OhLoRA/stylegan_and_segmentation/stylegan/synthesize_results_filtered

Related Link Full Dataset (10,000 images) Generated by StyleGAN model above

  3. Property Score Info

Property Score Detailed info

Property Score… See the full description on the dataset page: https://huggingface.co/datasets/daebakgazua/250408_OhLoRA_filtered_images.

minimal set of data and weights to be able to reproduce steps of the project

data/ -> individual samples + *.csv -> label files identifying samples

(3 sets, train / val / test and 3 dates per set with 45 lead times). 2160 samples total.

stats/ -> normalisation constants

weights/ -> checkpoint of the weights used in the article : StyleGAN networks and optimizers, betas + alphas (calibration), Covariance matrix for the whole latent space "whitening matrix"

outputs/ -> pre-created folder hierarchy to save results with the associated code. This is a simple indication and folder names/locations are flexible.

iFakeFaceDB is a face image dataset for the study of synthetic face manipulation detection, comprising about 87,000 synthetic face images generated by the Style-GAN model and transformed with the GANprintR approach. All images were aligned and resized to the size of 224 x 224.

1. Overview

All 10,000 images generated by GenForce StyleGAN celeba_partial-256x256, for Oh-LoRA Project.

  2. Save Path & Related Link

Save Path 2025_04_08_OhLoRA/stylegan_and_segmentation/stylegan/synthesize_results

Related Link Filtered Dataset (4,703 images | Subset of this dataset), filtered by CNN Models

  3. Gender & Quality Labels

Images Labels csv file

First 2,000 images both Gender & Quality (csv) All 10,000 images both Gender &… See the full description on the dataset page: https://huggingface.co/datasets/daebakgazua/250408_OhLoRA_all_generated_images.