News & Events

Nijmegen — Radboud University has launched SYNAPSIS, a collaborative project to help researchers in the social sciences and humanities share video and audio data while protecting the privacy of the people in those recordings. The project is funded by the Dutch Research Council (NWO) through its Thematic Digital Competence Centres programme.

Researchers across the social sciences and humanities increasingly rely on video and audio recordings to study human behaviour and social interaction — from classroom observations to clinical interviews and conversational data. Yet privacy concerns often prevent this valuable data from being shared or reused, limiting transparency and reproducibility in research.

SYNAPSIS addresses this challenge head-on. The project provides an open-source platform that integrates masking tools, allowing researchers to mask faces, mask voices, and remove other identifying features from recordings — all through a user-friendly interface that requires no programming expertise. The toolkit can be deployed on cloud infrastructure or set up locally. At the same time, the platform preserves research-relevant information such as body movement, gesture, and other behavioural signals.

"The goal is to make audiovisual research data as open as possible and as closed as necessary. Many researchers want to share their data but don't know how to do so responsibly. SYNAPSIS gives them the tools and the training."

Platform and training

SYNAPSIS has two main pillars. The first is a cloud-based platform built on the existing MaskAnyone toolkit and integrated with Dutch research infrastructure provider SURF. Researchers can upload recordings, apply masking, and archive privacy-compliant data — all within a secure, GDPR-compliant environment.

"Minimising privacy risks, maximising analytic utility in SSH — that is the core challenge SYNAPSIS is designed to address."

The second pillar is a comprehensive training programme — the Masking School — designed for researchers, data stewards, and support staff at Dutch universities. Workshops, hands-on sessions, and a summer school will build the digital skills needed to incorporate privacy-aware practices into everyday research workflows.

"Researchers and support staff tell us they find current masking tools too technical or hard to fit into their workflows. We are building something that is accessible to everyone, regardless of their technical background."

A collaborative effort

SYNAPSIS brings together a team of researchers from Radboud University, Vrije Universiteit Amsterdam, the University of Amsterdam, and Erasmus University Rotterdam, with support from infrastructure partners SURF and DANS, and international partner the Hasso Plattner Institute in Potsdam. The research infrastructure organisations CLARIAH and ODISSEI serve in an advisory capacity.

Over the next two years, the project will pilot the platform at several Dutch universities while training over 100 researchers and data stewards. The aim is to develop workflows and standards that can scale to other institutions in the Netherlands and beyond.

"Masking opens up new possibilities for open science, but it also raises important ethical questions. SYNAPSIS ensures these questions are part of the conversation from the start."

"Building sustainable research infrastructure means connecting the right tools with the right standards. SYNAPSIS bridges masking technology and established data archiving workflows."

Background

The project is funded under NWO's Thematic Digital Competence Centres for Social Sciences and Humanities (TDCC-SSH), which supports initiatives that strengthen digital skills and infrastructure in the SSH domain. SYNAPSIS responds to growing calls in the research community for better data management practices, including a recent policy shift arguing that audio and video recordings should not be destroyed by default but rather managed responsibly.

The SYNAPSIS platform builds on MaskAnyone, an open-source toolkit developed by members of the project team that enables face masking, voice anonymisation, and body kinematics extraction through a point-and-click interface.