PhD Position: Speech Privacy in Multi-Speaker and Multi-Modal Scenarios

1. Introduction

Are you fascinated by the human voice and what it reveals about the person speaking? Do you care about the privacy implications of speech data and supporting the development of responsible AI for speech? Are you interested in coming to the Netherlands for a PhD within a joint doctoral network? If so, we invite you to apply for this PhD position.

2. Job description

We are seeking a PhD candidate for the joint doctoral training programme on Privacy for Smart Speech Technology (PSST), which is funded by the Horizon Europe Marie Skłodowska-Curie Action. You will receive a joint doctorate from Radboud University and from the University of Lisbon. 

As a PhD candidate, you will conduct research on AI techniques for analysing and protecting personal information in speech recordings. Our voices contain information that reveals our identities and also other potentially privacy-sensitive information. Until now, research on speech privacy has focused on single speakers in audio-only scenarios, most often read speech. In this project, we move beyond this focus and closer to real-world scenarios, more specifically, to multi-speaker settings such as meetings or dialogues and also to multi-modal data, including text or video data.

Publishing such data, which is essential to R&D activities, may leak sensitive information, but this area is still under-researched. Specifically, the PhD project aims to identify revealing factors in multi-speaker and multi-modal data and find ways to detect and hide them using natural language processing and audio analysis. We look at anonymisation, but also go beyond to study attributes of speakers, which recent work has identified as an open issue.

The first 25 months of the PhD programme will be at Radboud University in Nijmegen and will be followed by 20 months at INESC-ID in Lisbon, and then by an additional 3 months at Radboud University. The PSST consortium includes 7 European universities and 11 organisations and companies that make practical use of speech privacy technology. An important collaboration partner for the PhD is ELDA, the Evaluations and Language resources Distribution Agency.

Your teaching load may be up to 10% of your working time.

3. Profile

• You have a Master’s degree in electrical engineering, computer science or a related area relevant to the position.
• You have a background in speech processing. An affinity with NLP and/or computer vision would also be welcome.
• You have experience with empirical research in speech technology, evaluating AI systems, and socio-technical factors of AI.
• You have an understanding of and appreciation for the importance of data protection and privacy.
• You have solid programming and software engineering skills.
• You have an excellent command of written and spoken English.
• You are proactive and enjoy engaging with others during network-wide training events, international mobility, and public dissemination activities.  

You have not resided or carried out your main activity (work, studies, etc.) in the Netherlands for more than 12 months in the 36 months immediately before your recruitment date. A Master’s degree must have been completed before employment can start. You must not already hold a PhD.

4. We are

As a PhD candidate, you will be affiliated with the Data Science Group at the Institute for Computing and Information Sciences (iCIS), part of the Faculty of Science at Radboud University. The Data Science Group comprises around 50 researchers with expertise in machine learning, information retrieval and other specialisations that are important for working with data. The group is involved with a large number of projects from medical to industrial, including projects that are developing new techniques in machine learning and work with large language models. You will collaborate closely with researchers at the Centre for Language Studies at the Faculty of Arts, who have expertise in speech science, speech recognition and NLP.

5. We offer

• We will give you a temporary employment contract (1.0 FTE) of 1.5 years, after which your performance will be evaluated. If the evaluation is positive, your contract will be extended by 2.5 years (4-year contract). During the time you spend in Portugal, you will have a separate contract with INESC-ID.
• You will receive a starting salary of €3,059 gross per month based on a 38-hour working week, which will increase to €3,881 in the fourth year (salary scale P).
• You will receive an 8% holiday allowance and an 8,3% end-of-year bonus. 
• We offer Dual Career Coaching. The Dual Career Coaching assists your partner via support, tools, and resources to improve their chances of independently finding employment in the Netherlands. 
• You will receive extra days off. With full-time employment, you can choose between 30 or 41 days of annual leave instead of the statutory 20. 

6. Practical information and applying

You can apply only via the button below. Address your letter of application to Martha Larson. In the application form, you will find which documents you need to include with your application.
Note that in order to apply, you must not have resided or carried out your main activity (work, studies, etc.) in the Netherlands for more than 12 months in the 36 months immediately before your recruitment date.

The first interviews will take place on 9 and 10 October. You will preferably start your employment as soon as possible.

We can imagine you're curious about our application procedure. It describes what you can expect during the application procedure and how we handle your personal data and internal and external candidates.