Speech is one of the most expressive and effortless ways that we communicate with one another. For those who are severely paralyzed and also cannot speak verbally, it can be difficult to communicate their needs and express themselves, decreasing their quality of life and independence. With recent advances in machine learning and neural recording hardware, a speech brain computer interface (BCI) could restore natural, effortless communicative abilities to these individuals.
Part of my PhD work focuses on improving and developing speech BCIs. I work on developing speech production decoding algorithms using electrocorticography (ECoG) and computational approaches, like signal processing and machine learning. I also work on more basic science research concerning speech production.
I have a particular interest in deep learning, unsupervised learning, and the neuroethics involved in our research.