Kristian Georgiev's repositories
Wireless-3D-pose-tracking
Advanced systems have been developed to enable accurate room-scale 3D tracking for virtual reality applications. Steam has recently made these tools available for community development (https://partner.steamgames.com/vrlicensing, https://www.triadsemi.com/steamvr-tracking/). These tools can be leveraged to enable wireless 6-degree of freedom tracking of freely moving animals for experimental neuroscience. This would provide a large number of benefits compared to current video-based tracking technology and would likely become the _de-facto gold-standard method in the field, once implemented. A functional system requires the development of calibration and sensor fusion algorithms to combine 15-60Hz room-scale optical tracking with 500Hz local inertial reference frame data. This UROP would focus on the development and verification of these algorithms in a high level programming language and then implementing them on existing head-borne FPGAs and microcontrollers being used in the Wilson lab for behavioral research.
whocreatedme
Automate file origin tracking in your projects
privacy-induces-robustness
Code for "Privacy Induces Robustness: Information-Computation Gaps and Sparse Mean Estimation"
awesome-equivariant-network
Paper list for equivariant neural network
ffcv
FFCV: Fast Forward Computer Vision (and other ML workloads!)
HowToTrainYourMAMLPytorch
The original code for the paper "How to train your MAML" along with a replication of the original "Model Agnostic Meta Learning" (MAML) paper in Pytorch.
implicit-bias-of-linear-equivariant-networks
Code for Implicit Bias of Linear Equivariant Networks
kristian-georgiev.github.io
Personal website
learn2learn
A PyTorch Library for Meta-learning Research
Perfect-Matchings
A script that generates a random biadjacency matrix for a d-regular bipartite graph in linear time and computes the parity of its determinant.
robustness_rot
A library for experimenting with, training and evaluating neural networks, with a focus on adversarial robustness.
setup.py
📦 A Human's Ultimate Guide to setup.py.
stopwords-de
German stopwords collection
TRAK-1
Code for TRAK: Understanding Model Predictions at Scale