A speech annotation system developed for English that identifies landmarks and other acoustic cues to distinctive features (Stevens 2002C, Huilgol et al. 2019) will be expanded to many languages. I am working towards generalizing the annotation algorithm so that if given a matrix mapping the phonemes to the different distinctive features of a language, the annotation algorithm will correctly generate the predicted standard cue labels for the speech file. Naturally, each language will have its own set of linguistic rules and exceptions that do not directly map to other languages, but this research aims to create the ground work for a flexible annotation algorithm and set of rules that generalize to many languages. For example, Spanish has 14 diphthongs while English has three diphthongs.

I want to delve deeper into my research. At the beginning of my research, I didn’t have much knowledge about researching or the tools/methods used. However, with time spent researching and taking classes, I am at a much better place knowledge wise than I was before. I want to learn how to effectively communicate my research results to a diverse audience. I am excited about the challenges I may encounter in the process and about overcoming them.