PhD student position in Generative Modelling of Conversational Dynamics at the School of Electrical Engineering and Computer Science at KTH Royal Institute of Technology.
Project description
This project aims to create generative models of spoken conversation that enable speaking machines to adapt their conversation style over time, in the same natural manner that we humans do. To make this a reality, you will perform self-supervised learning of aspects such as speech patterns, rhythm, and intonation from human dialogues, and use the results to build better, adaptive text-to-speech systems. This represents a new way to integrate text-to-speech technology, dialogue systems, and generative machine learning. By being speaker-agnostic, the approach seeks to do this without propagating existing biases, to create engaging and inclusive human-computer interactions.
In order to succeed as a doctoral student at KTH you need to be goal oriented and persevering in your work. During the selection process, candidates will be assessed upon their ability to:
- independently pursue their work
- collaborate with others,
- have a professional approach and
- analyse and work with complex issues.
Other important traits are curiosity, attention to detail, and openness toward interdisciplinary research.
The applicant needs to demonstrate excellent command of deep learning, including strong programming skills in PyTorch or similar. Experience in areas like speech technology, large language models (LLMs), self-supervised learning, and signal processing is also desirable.
After the qualification requirements, great emphasis will be placed on personal competency.