How much private information can a single spoken word inadvertently reveal?

The rapid growth of smart speech technology is transforming how people access information and applications through the human voice. However, these advancements come with privacy risks. Smart devices like speakers and smartwatches can expose users to the exploitation of sensitive information within speech signals, such as one’s health status, without their knowledge or consent.

Can innovative privacy-preserving smart speech technologies reduce these risks? Aligned with this mission and its ongoing commitment to strengthen the internationalisation of its advanced training, INESC-ID has recently joined the Marie Curie doctoral training network “Privacy for Smart Speech Technology” (PSST). The PSST consortium includes world-leading research labs in speech processing, privacy, and usability, along with numerous innovative industrial partners across Europe.

The network will train a new group of researchers to develop these urgently-needed technologies using cutting-edge deep learning methods. This double-degree, high-impact training program partners with Aalto University (Finland, Coordinator), INESC-ID (Portugal), EURECOM GIE (France), INRIA (France), Ruhr-Universität Bochum (Germany), Radboud University Nijmegen (the Netherlands), and the Technical University of Berlin (Germany), among other partners. It will provide doctoral researchers with expertise in speech processing, machine learning, and comprehensive knowledge of privacy threats and privacy-enhancing technologies.

Funded with a total of €4M by the Marie Skłodowska-Curie Actions (MSCA) Doctoral Networks (DN) program, the program will run for 60 months, starting January 25, 2025. INESC-ID’s participation will be led by Isabel Trancoso, an INESC-ID researcher in the area of Human Language Technologies and full professor at Instituto Superior Técnico, with funding of €487K.

PSST partners and researchers will jointly create a unique, user-centred view of smart speech technologies, reducing threats like exploitation of sensitive information via speech. They will also promote innovative business models for responsible and sustainable privacy-aware technologies in the EU.

The call for open positions begins in Fall 2024.

Click here for more detailed information and full list of partners.