emaerthin/microevolution-lang-phones

This repo stores various ideas and approaches to the microevolution of individual speaker's phonemes retrieved from recordings found on youtube

microevolution-phones

This repo stores ideas and approaches to the microevolution of individual speaker's phonemes retrieved from recordings found on youtube

Set-up

The project is written in python. It requires setting the environment using Pipenv. First ensure that pipenv is already installed - if not, type pip install pipenv.

Now, type pipenv install and pipenv shell et voila - you should be able to run scripts seamlessly now.

Third party

• Multicore-TSNE
• pocketsphinx
• pytube

Multicore-TSNE

This is a multicore modification of Barnes-Hut t-SNE by L. Van der Maaten with python and Torch CFFI-based wrappers. This code also works faster than sklearn.TSNE on single core.

Officially distributed library has a defect that it occupies just a single core when run on macOS. Hence, sources from macOS complaint fork are used: https://github.com/sg-s/Multicore-TSNE

(see Issue #53: Using single core even when n_jobs=4 is used)

pocketsphinx

PocketSphinx is a lightweight speech recognition engine, specifically tuned for handheld and mobile devices, though it works equally well on the desktop.

While an official distribution of pocketsphinx is installed with pipenv, source repository contains an officially distributed generic US english acoustic model trained with latest sphinxtrain.

pytube

A lightweight, dependency-free Python library (and command-line utility) for downloading YouTube Videos. https://python-pytube.readthedocs.io

Sometimes (like as of 2019-04-27) official distribution of pytube is not working because youtube API changed in the meantime. This will be patched officially for sure at some day but until then a forked repository with local fix is used.