Hannes Stärk's Projects
Making self-supervised learning work on molecules by using their 3D geometry to pre-train GNNs. Implemented in DGL and Pytorch Geometric.
Unsupervised method for binding site prediction using attention patterns of protein language models.
Removing background noise from clips of speech and improving audio quality (PyTorch)
TensorFlow code and LaTex for Bachelor Thesis: Understanding Variational Autoencoders' Latent Representations of Remote Sensing Images :earth_africa:
Get protein embeddings from protein sequences
R code for NetworkCentralityCalculator. A web-tool with 5 different centrality measures. LaTex and pdf for documentation and explanation of different measures with a focus on "dependency centrality".
A library for graph deep learning research
A modular framework for neural networks with Euclidean symmetry
Using similarity in embedding space for predicting EC numbers
EquiBind: geometric deep learning for fast predictions of the 3D structure in which a small molecule binds to a protein
Implementation of FlowSite and HarmonicFlow from the paper "Harmonic Self-Conditioned Flow Matching for Multi-Ligand Docking and Binding Site Design"
Representing robots as graphs for reinforcement-learning in PyBullet locomotion environments.
Code for my website built with Angular and running on GitHub Pages.
Pre-train and evaluate Graph Neural Networks or Transformers on molecules with the ELECTRA method.
Benchmark datasets, data loaders, and evaluators for graph machine learning
Using Transformer protein embeddings with a linear attention mechanism to make SOTA de-novo predictions for the subcellular location of proteins :microscope:
code for the SE3 Transformers paper: https://arxiv.org/abs/2006.10503
Embed human pose information into neural radiance fields (NeRF) to render images of humans in desired poses :running: from novel views