mkurop Goto Github PK

aberth-method

Aberth Method, a root-finding algorithms for simultaneous approximation of all the roots of a univariate polynomial.

audiofile

A simple C++ library for reading and writing audio files.

boxes3d

Spliting an arrangement of 3D boxes into disjoint boxes

composite-measure

Composite measure of speech qualilty from the book by Philipos Loizou "Speech Enhancement - Theory and Practice"

counting_monomials

Counting monomials with even exponents in a trace of a matrix raised to power

cuda-metaprog

Illustration of CUDA expression templates/meta programming

cut_wave_with_speech_into_shorter_fragments

Cutting a wave file containing speech into shorter fragments of required length. Uses speech pauses as cut points.

encryption_decryption_bash_scripts

Encryption and decryption using openssl with ssh-keygen generated rsa key pair. Uses hybrid cryptography.

gmms

Classical Expectation-Maximization based Gaussian mixtures models training on CPU or GPU.

junctiontree

Algorithm for turning bayesian graph or factor graph into junction tree.

kaldi

kaldi-asr/kaldi is the official location of the Kaldi project.

kmeans

KMeans algorithm for computing a codebook based on training set - uses NumPy or CuPy.

kmeanspp

Functions for running kmeans++ algorithm, either on CPU or GPU

levenshtein_distance

Class for computing the Levenshtein distance and the three kinds of errors: insertions, deletions, and substitutions

mkurop.github.io

nisqa

NISQA - Non-Intrusive Speech Quality and TTS Naturalness Assessment

noise-tracking

Fast nonstationary noise tracking by Q. Zhang et al.

noise-tracking-hendriks

Noise tracking algorithm developed by Richard Hendriks

noise-tracking-matlab

Fast nonstationary noise tracking algorithm by Q. Zhang et al. - coded in Matlab/Octave.

p56

Implementation of the ITU P.56 active speech RMS power and active speech factor measurement.

packs

Memory efficient tables of arbitrary size numbers from a given range.

spacevim_python_config

My SpaceVim configuration init.toml file for turning it into Python IDE.

spsymp2019

Documentation of the algorithm for finding a region of the high probability of a given 2-dimensional pdf.

stpkalman

Extended Kalman filter for denoising autoregressive (AR) proces, e.g. speech, contaminated by noise, also an AR process.

stpltpdecoder

Classes for performing analysis and synthesis based on the Short Term Predictor/Long Term Predictor (STP/LTP) speech signal model. This speech signal model originates from the speech coding paradigm known as the Code Excited Linear Prediction (CELP).

vad

Voice activity detection based on Sohn's algorithm with noise tracking by Zhang.

vad-python

Voice activity detection by Sohn coded in python.