Postdoc in speech processing and speech technology

Updated: about 2 years ago
Deadline: 07 Jan 2022

KTH Royal Institute of Technology in Stockholm has grown to become one of Europe’s leading technical and engineering universities, as well as a key centre of intellectual talent and innovation. We are Sweden’s largest technical research and learning institution and home to students, researchers and faculty from around the world. Our research and education covers a wide area including natural sciences and all branches of engineering, as well as architecture, industrial management, urban planning, history and philosophy.


Job description

We are seeking a postdoc for the project “Multimodal encoding of prosodic prominence in voiced and whispered speech” led by dr Zofia Malisz at KTH Royal Institute of Technology in Stockholm, Sweden. The project is part of our larger goal to understand the speech signal and speech communication behaviours via careful analysis as well as modelling in technical systems. As such, the goal is aligned with the wider objectives of digital linguistics where advanced computational and machine learning methods are applied to test linguistic hypotheses.

In particular, the project seeks to understand the differences in how  information is encoded in whispered vs. modally voiced speech. We focus on the acoustics of prosodic encoding and use Information Theory as an explanatory framework. In the course of the project, we collect parallel whispered and voiced databases and process them using signal processing and machine learning techniques. We then extract information about the acoustic, prosodic (as well as gestural) variability and analyse the impact of information theoretic predictors (such as linguistic surprisal) on the variability. Finally, we train and evaluate speech synthesis models capable of whisper-to-speech conversion and prosodic control based on linguistic surprisal.

This work builds on results regarding the analysis of information encoding of prosodic prominence in voiced and whispered speech as well as prosodic control in speech synthesis (Malisz, Brandt, Möbius, Oh, Andreeva 2018; Malisz, Jonell, Beskow 2019; Döhler Beck, Wennberg, Henter, Malisz 2021).   

Your work will involve the preparation of speech processing pipelines capable of analysing, processing and synthesising whispered speech. You will also be involved in statistical modelling of the gathered data and updating hypotheses regarding the data within the framework of Information Theory. The project is highly interdisciplinary, and we would like to see applicants with interests in both signal processing, speech sciences, and machine learning.

The work will take place in the speech group at the department of Speech Music and Hearing, at KTH – an internationally recognized research lab in speech and language technology. The research at the department is focused on understanding human- and human-machine communication based on multimodal information. This research area is truly multi-disciplinary, bridging computer science, machine learning, linguistics, and perception and cognitive disciplines. For more information, see https://www.kth.se/is/tmh


What we offer
  • A position at a leading technical university that generates knowledge and skills for a sustainable future
  • Engaged and ambitious colleagues along with a creative, international and dynamic working environment
  • Works in Stockholm, in close proximity to nature
  • Help to relocate and be settled in Sweden and at KTH
  • Work in an interdisciplinary team of speech engineers, linguists and machine learners that will enable you to develop unique competences

Read more about what it is like to work at KTH


Qualifications

Requirements

A doctoral degree or an equivalent foreign degree, obtained within the last three years prior to the application deadline (With some exceptions for special reasons such as periods of sick or parental leave, kindly indicate if such reason exists in your resume). Applicants should have a doctoral degree in a subject relevant for the research, such as signal processing, speech technology, machine learning or computational modeling for speech applications or speech sciences. The position requires good skills in signal processing and knowledge of speech science and acoustics. Skills in machine learning and statistical analysis will be preferred. Good command of English, in writing and speaking, is a prerequisite for presenting research results in international periodicals and at conferences. We also expect applicants to have a deep interest in speech sciences and technology, and to have done their PhD in a related area. 

Preferred qualifications

Candidates with interdisciplinary backgrounds are especially welcome to apply. We would like to see that you have experience of formulating research ideas independently from your supervisor and getting them implemented. At the same time, experience of doing research in collaboration with others is commendable. You should also be aware of diversity and equal opportunity issues. 

Great emphasis will be placed on personal competency


Trade union representatives

You will find contact information to trade union representatives at KTH's webbpage .


Application

Log into KTH's recruitment system in order to apply to this position. You are the main responsible to ensure that your application is complete according to the ad.

The application must include:

  • CV including relevant professional experience and knowledge.
  • Copy of diplomas and grades from your previous university studies. Translations into English or Swedish if the original documents have not been issued in any of these languages.
  • Brief account of why you want to conduct research, your academic interests and how they relate to your previous studies and future goals. Max two pages long.
  • One or two letters of recommendation

Your complete application must be received at KTH no later than the last day of application, midnight CET/CEST (Central European Time/Central European Summer Time).


About the employment

The position offered is for, at the most, two years.

A position as a postdoctoral fellow is a time-limited qualified appointment focusing mainly on research, intended as a first career step after a dissertation.


Others

Striving towards gender equality, diversity and equal conditions is both a question of quality for KTH and a given part of our values.

For information about processing of personal data in the recruitment process please read here.

We firmly decline all contact with staffing and recruitment agencies and job ad salespersons.

Disclaimer: In case of discrepancy between the Swedish original and the English translation of the job announcement, the Swedish version takes precedence.



Similar Positions