Astrid Frykman

I am a voice science engineer at the Division of Speech, Music and Hearing (TMH) at KTH. I apply representation learning to voice maps built from microphone and electroglottograph (EGG) recordings. My focus is on understanding the nonlinear aspects of voice production, mainly through variational autoencoders.

I hold a Master of Science in Engineering (Civilingenjör) in Engineering Physics from KTH, with a specialization in machine learning.

Before that, I was a competitive sailor and represented Canada at several international regattas, including the Youth Sailing World Championships and the 29er Open World Championship, where my team placed second among female teams in 2017.

Outside research, I practise Vipassana meditation daily and teach yoga at the Student Sangha Circle, a THS association I founded.