GAN Clothes Generator

Overview

Welcome to the GAN Clothes Generator project! This project aims to generate high-quality images of clothes using Generative Adversarial Networks (GANs). By training a GAN on a dataset of clothing images, we can create realistic and diverse new clothing designs.

Table of Contents

• Introduction
• Features
• Installation
• Usage
• Model Architecture
• Dataset
• Results
• Contributing
• License

Introduction

This project leverages the power of GANs to generate novel images of clothes. The main components include a generator that creates new images and a discriminator that evaluates their realism. By iterating through training, the generator improves its ability to produce realistic clothing images.

Features

• High-Quality Image Generation: Generates realistic and diverse clothing images.
• Customizable Training: Easily modify training parameters and architecture.
• Pre-trained Models: Includes pre-trained models for quick setup.

Model Architecture

• The model has a combination of wide and narrow network, making it efficient on sub-structures of the features and on overall image.

Dataset

• The model is trained to generate images similar to FashionMNIST data. It generates grayscale images of 28 x 28 size.

Installation

Clone the repository and install the required dependencies.

git clone https://github.com/yourusername/gan-clothes-generator.git
cd gan-clothes-generator
pip install -r requirements.txt