This projects were made for the Machine Learning course at the AGH UST in 2024/2025. Obtained maximum grade 5.0.

Testing and playing around with VAEs using the mnist dataset

These are the results for Focal Frequency Loss for Image Reconstruction and Synthesis paper but with different dataset. In this case the TFW: Annotated Thermal Faces in the Wild Dataset. By using novel focal frequency loss it allows a model to adaptively focus on frequency components that are hard to synthesize by down weighting the easy ones.

archived content from probabilistic learning course

Stable-Diffusion-Core is a complete implementation of the Stable Diffusion model from scratch. This repository includes the core components necessary for training and deploying diffusion models, including encoders, decoders, CLIP encoding, and U-Net architectures.

Deep Generative Neural Networks

Deep Learning for Computer Vision

Reconstructing images using VAE

Using variation autoencoder on multiple datasets. Started with simple Pokemon

Collection of materials for teaching about autoencoders

Quick and dirty test of a VAE on MNIST data.

Implemented the Variational AutoEnoder and Decoder

제 12회 산업통상자원부 공공데이터 활용 아이디어 공모전 최우수상 | LLM | 추천시스템

Official implementation of "DiVERT: Distractor Generation with Variational Errors Represented as Text for Math Multiple-choice Questions" (EMNLP 2024)

This repository contains PyTorch implementations of modern generative models, with extensive experiments on cat image generation.

A simple variational autoencoder

Research Internship - Institute for Information Processing, Leibniz University Hannover

Сross Domain Autoencoding Variational Bayes Algorithm on MNIST

Experiments with VAEs on MNIST dataset using Flux.jl