cgan

Experimental image data generation for boosting plant disease detection.

Ground motion simulation in the time-frequency domain by conditional GAN and phase retrieval

AI-powered text-to-image generation system using Stable Diffusion, Conditional GAN (CGAN), Hugging Face Transformers, and Cross-Attention mechanisms.

This project implements a Text to Image Generator using a Conditional Generative Adversarial Network (GAN) for synthesizing floorplan images from textual descriptions. It includes features such as custom dataset handling, performance metrics like FID and IS, and configurable training options for optimization.

CGAN and DCGAN networks used to generate card suits

The objective of this project was to develop a generic neural network model that, once trained, can generate new, non-existent textures. While initially designed for use in Blender, the model can be adapted for other software, including non-graphic programs, with appropriate adjustments.

NebulAI is a web app that generates realistic galaxy images using a Conditional GAN (cGAN) guided by K-Means clustering. Users can explore and generate galaxies based on learned visual categories through an interactive React + Django interface.

A PyTorch implementation of Conditional GAN (cGAN) to generate specific fashion category images (e.g., sneakers, bags) using FashionMNIST. Features custom Generator/Discriminator architecture with label embeddings, visual validation pipeline, and quantitative evaluation using FID score.

Astronomical image super-resolution using a pix2pix-based CGAN to translate Euclid galaxy images into JWST-like images.

Automatic Colorization of images using cGAN

Python code to convert hand drawn sketches into abstract art using GANs

Manga Colorization and Style Transfer

GAN, and cGAN models based on CIFAR10 dataset using Torch library.

심화 R&E

Implementation of a cGAN to perform a data generation task on the CIFAR10 dataset. This generation can be used both to augment the original CIFAR10 dataset or to generate a new dataset, based on CIFAR10 classes, from scratch. It also has been implemented a pre-trained classifier in order to evaluate the performance of the cGAN model.

SyntheticOcean: Open-Source Library for Generating Synthetic Tabular Data + SynDataGen (Framework for Synthetic Data Generation)

Domain adaptation on gaze estimation from MPIIFaceGaze to ColumbiaGaze

🎨Implementations of colorization algorithms.

CGAN for vehicle trajectories prediction

Repository for (Un)Clear SoC Project, done in the Summer of 2021.