Josh Meyer’s Website
Excerpt
Hi! My name’s Josh and I work on Automatic Speech Recognition, Text-to-Speech, NLP, and Machine Learning. This blog is some of what I’m learning along the way. All opinions are my own.
Speech Recognition
-
Mar 21, 2020
An Overview of Multi-Task Learning in Speech Recognition
-
Aug 17, 2019
My INTERSPEECH Schedule
-
Aug 17, 2019
Kaldi Troubleshooting Head-to-Toe
-
Aug 17, 2019
Kaldi Hyperparameter Cheatsheet
-
Nov 9, 2017
Kaldi nnet3 notes
-
Oct 13, 2017
Kaldi on AWS
-
Sep 29, 2017
Josh’s Speaker ID Challenge
-
Apr 5, 2017
Seminal Papers in ASR
-
Jan 10, 2017
How to use an Existing DNN Recognizer for Decoding in Kaldi
-
Dec 15, 2016
How to Visualize a Word Lattice with Kaldi
-
Dec 15, 2016
How to Train a Deep Neural Net Acoustic Model with Kaldi
-
Sep 12, 2016
How to use an Existing GMM Recognizer for Decoding in Kaldi
-
Feb 1, 2016
Some Kaldi Notes
-
Jan 27, 2016
CMU-Sphinx Cheatsheet
-
Jan 26, 2016
Installing Kaldi
-
Jan 9, 2016
The CMU-Sphinx Speech Recognition Toolkit: First Steps
Speech Synthesis
Machine Learning
Miscellaneous
Downloads
You can download an NVDA installer with Kyrgyz language support here. You should be able to install the program by double clicking on the file and following the directions. To turn Kyrgyz language support on or off, navigate to “Voices” under “Settings” after you’ve installed the program. This project was conducted with Empower Blind People, a non-profit organization for blind people in Kyrgyzstan. Any feedback on the Kyrgyz support (accent, translation, errors, etc) is gladly welcomed!
Lectures & Talks
Practical AI 104: Speech tech and Common Voice at Mozilla – Listen on Changelog.com
News about the Hakha Chin language being added to Mozilla’s Common Voice. The project was spear-headed by Peng Hlei Thang and the Linguistics Department at Indiana University Bloomington.
Interview during the Week of Young International Scientific Talents (Semaine des jeunes talents scientifiques internationaux):
Here’s another interview from the same week at France Inter.
Here’s a couple videos below about our speech synthesis project for the Kyrgyz language. This project was done in collaboration with Empower Blind People to create a speech synthesizer for the Kyrgyz language, to be used in the open source project NVDA.