KikiOduro/Ai_finalproject_AfricanFabricStyleGAN

Creating quality batik images from african print images. # Group26_AfricanFabricStyleGAN-Batiq 🌍 Generating African Batik Prints Using StyleGAN Welcome to the Batiq project! This repository contains the code and resources for generating African Batik prints using StyleGAN. We aim to leverage the power of deep learning to create unique and authentic Batik designs, celebrating the rich heritage of African culture. ## Project Overview 💡 This project aims to develop a machine-learning model to generate African Batik prints using StyleGAN. The primary goal is to produce unique Batik designs that artisans and designers can use. The project involves data collection, model development, validation, testing, and creating a user-friendly interface for generating new prints. ## Dataset 📊 We used a kaggle dataset containing images of African Batik prints. We preprocess the data by resizing images and normalizing pixel values. Link to Kaggle dataset here ## Data Preprocessing 📋 The data preprocessing involves: - Resizing images to a consistent size for model input. - Normalizing pixel values to the range [-1, 1]. ## Feature Engineering 🔍 Feature engineering steps include: - Extracting style and content features from the images. - Using AdaIN (Adaptive Instance Normalization) to combine content and style features. ## Model Training 🧑🏻‍💻 The StyleGAN model is built using TensorFlow and Keras. We trained the model on the dataset, adjusting hyperparameters such as the number of layers, the number of neurons per layer, and the learning rate. We used techniques like regularization and dropout to prevent overfitting. ## Evaluation and Optimization 📈 Once the model was trained, we evaluated its performance using various metrics. We computed metrics such as Frechet Inception Distance (FID) to assess the quality of generated images. Optimization techniques, including parameter tuning and cross-validation, were employed to enhance model performance. ## Usage 💪🏻 To run the project locally, follow these steps: 1. Clone this repository: `` …