This repository contains the code, data and scripts used to write the Bachelor Thesis "Latent representations for traditional music analysis and generation".
-
Updated
Jul 25, 2022 - Python
This repository contains the code, data and scripts used to write the Bachelor Thesis "Latent representations for traditional music analysis and generation".
Toy example for a Conditional Variational Autoencoder in Keras. Regresses features from automatically generated images. Useful for learning about the concept.
Code used in blog post about dimensionality reduction using Python
Exploring the use of Adversarial Constrained Autoencoder Interpolation (ACAI) to improve the quality of latent space for 3D human pose representation using the h36m dataset.
Project for the course "Deep Learning" - UniTS
Extension of GANspace: https://github.com/harskish/ganspace
A vector database for images, with the possibility to run a similarity search. I implemented this as I needed to make a project with GPT4o vision API
TensorFlow implementation of a simple Variational Autoencoder. Includes a simple GUI for exploring the learned latent space.
Multi-Operational Mathematical Derivations in Latent Space
Playing wiht Latent Space of VAE with PSO
Witchcraft is a toolkit capable of encoding documents of various content types and structure (flat, hierarchical, or series) into searchable latent vector space.
Pipeline Consisting of LSTM + Variational and Transformer Based Autoencoders + PCA/UMAP (Parameterized and Non-Parameterized) For Generating Low-Dim Manifold Representation of V1 Neural Activity
Replication of the research paper titled Auto-Encoding Variational Bayes.
Supplementary material for the paper "Lightweight Multitask Learning for Robust JND Prediction using Latent Space and Reconstructed Frames", IEEE TCSVT, 2024.
A project for understanding latent spaces in different neural networks (joint work with interns 2018)
A playground for image generation using diffusion models
Modeling defocus blur with linearity constraints in the latent space
VecDB is a tool to help you to store, manage and compare embeddings vectors for machine learning applications.
Code implementation of the detection network capable of dealing with many overlapping spline bodies.
Add a description, image, and links to the latent-space topic page so that developers can more easily learn about it.
To associate your repository with the latent-space topic, visit your repo's landing page and select "manage topics."