Writing

Notes from the work

Ramblings about AI, speech, and language: how the systems actually work, what breaks once they meet real data, and the occasional detour into how we got here. Written while figuring things out, so some of it is a lab notebook and some of it is an argument.

Subscribe via RSS

Arabic ASR and DI (Red Hen Lab - GSoC 2018)

In this post, we will summarize the work done on Arabic speech recognition and Arabic dialect identification projects for RedHenLab as part of GSoC 2018. Each of…

Week 8: Gale Arabic on Case HPC

It's week 8 of my GSoC project with Red Hen Lab. As mentioned in an earlier post, running the system on Case HPC is one of the main goals of this GSoC project. I…

Week 7 (Cont.): Arabic Speech Data

Some licensing issues related to the MGB-2 data set resulted in the data not being available for a while (I wish I will be able to work with it before the end of…

Week 7: Dialect Identification

Sixth week coding my Google Summer of Code project with Red Hen Lab. I decided to use the time to go into the dialect identification task while the training…

Week 5: Case HPC

Hello! This is my fifth week coding my Google Summer of Code project with Red Hen Lab. One of the goals of this project is to implement the code on Case HPC,…

Week 4: Acoustic Modeling

Fourth week coding my Google Summer of Code project with Red Hen Lab. Obtaining the data turned out to be a bit tricker than I thought (due to some license…

Week 1: Language Modeling

This is my first week coding my Google Summer of Code project with Red Hen Lab. My first task will be preparing the language model to use for decoding.