dc-gan

My implementation of various GAN (generative adversarial networks) architectures like vanilla GAN (Goodfellow et al.), cGAN (Mirza et al.), DCGAN (Radford et al.), etc.

Train a DCGAN model on Colaboratory to generate Steam banners.

Generation of Human-Like handwritten digits using different GAN Architectures. The models were developed using Low-Level Tensorflow.

Implemented Vanilla RNN and LSTM networks, combined these with pretrained VGG-16 on ImageNet to build image captioning models on Microsoft COCO dataset. Explored use of image gradients for generating new images and techniques used are Saliency Maps, Fooling Images and Class Visualization. Implemented image Style Transfer technique from 'Image Style Transfer Using Convolutional Neural Networks'. Implemented and trained GAN, LS-GAN and DC-GAN on MNIST dataset to produce images that resemble samples from MNIST, DC-GAN gave best resembling images.

Implementations of GANs in PyTorch for Pokemon image generation

A very simple and plain DC GAN to generate Image Flower Pictures out of the dataset.

Generation of images from NORB dataset using DC-GAN

This project explores the use of Generative Adversarial Networks (GANs) including DCGAN, ACGAN, and ProGAN to create realistic synthetic chest X-ray images. It aims to enhance medical imaging datasets and improve the performance of diagnostic AI models, especially in data-scarce scenarios.

Deep Learning model that generates Pokemon images

Implement Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks(DCGAN)

Generating fake images using DC-GANs

DC Generative Adversarial Network for an MNIST Handwritten Digits From Scratch in Keras & tensorflow.

A DCGAN in TensorFlow/Keras to generate artificial human faces, featuring an interactive web UI built with Streamlit for easy inference. This project was developed as part of a winter internship at IIT Guwahati.