About Me

<aside> <img src="/icons/user-circle_orange.svg" alt="/icons/user-circle_orange.svg" width="40px" /> Hey! I’m Atli and I work on speech generation models! I’m interested in controllability and how we interact with complex generative models. I’m currently wrapping up my PhD at CSTR, University of Edinburgh.

</aside>

<aside> <img src="/icons/link_orange.svg" alt="/icons/link_orange.svg" width="40px" /> Google scholar

</aside>

<aside> <img src="/icons/link_orange.svg" alt="/icons/link_orange.svg" width="40px" /> LinkedIn

</aside>

<aside> <img src="/icons/link_orange.svg" alt="/icons/link_orange.svg" width="40px" />

CV

</aside>

<aside> <img src="/icons/link_orange.svg" alt="/icons/link_orange.svg" width="40px" /> Email

</aside>

At the moment

Posts

<aside> <img src="/icons/clock-alternate_orange.svg" alt="/icons/clock-alternate_orange.svg" width="40px" /> Work in progress

</aside>

Papers I’ve written

This is a list of papers I’ve published as part of my PhD as well as work I conducted as part of the Icelandic language technology development program.

<aside> <img src="/icons/info-alternate_orange.svg" alt="/icons/info-alternate_orange.svg" width="40px" /> Click the arrow to see further information about each paper

</aside>

RepeaTTS: Towards Feature Discovery through Repeated Fine-Tuning

Just because we camp, doesn’t mean we should: The ethics of modelling queer voices

A Human-in-the-Loop Approach to Improving Cross-Text Prosody Transfer

Controllable Speaking Styles Using A Large Language Model

Do Prosody Transfer Models Transfer Prosody?

Talrómur: A large Icelandic TTS corpus

Manual speech synthesis data acquisition-from script design to recording speech

Other resources

This list contains Icelandic speech corpora I’ve worked, corpora creation tools and TTS models.