FashionCraft is a project that stems from my deep passion for fashion. It's driven by the desire to explore the intersection of creativity and technology, blending the elegance of traditional fashion with the boundless possibilities offered by AI. My aspiration is to push the boundaries of conventional fashion, utilizing machine learning to showcase the beauty and diversity of clothing.
In this project, I employed Generative Adversarial Networks (GANs) to generate realistic clothing images.
I utilized the Fashion MNIST dataset, which consists of grayscale images of various fashion items. TensorFlow was employed to automatically download and preprocess this dataset.
The Generative Adversarial Networks (GANs) contain:
-
Generator network: This network takes random noise as input and generates synthetic fashion images. It consists of upsampling layers, convolutional layers, and activation functions.
-
Discriminator network: This network helps to distinguish between real and fake fashion images. It consists of convolutional layers, LeakyReLU activations, and dropout layers.