Speech Tech Lab

About Us

The Speech Technology Lab specializes in speech technology challenges, with special emphasis on “small data” and social impact. We want to see what can be done with as little data as possible for the greater good. Consequently, we work on topics around small languages, improved human-machine interaction, and speech tech ethics. A cornerstone of our work is our commitment to co-development with local partners, ensuring that our technological solutions are culturally appropriate and ethically sound. Our interdisciplinary team combines technical prowess in speech technology with a deep understanding of the ethical implications of AI in linguistic contexts. We're well-positioned to contribute to projects that bridge the digital divide in language technology.

Example projects

Below are projects which exemplify the work we do in the Speech Tech Lab.

• Frisian Synthesis: We developed a synthetic voice for Frisian, which can be used freely for non-commercial purposes. The techniques developed in this initiative extend to other small languages. Some outputs: [1], [2], [3]; in the media: [1], [2], [3]. (Lead: Phat Do) 

• Melodies of the North: We curate and develop resources to propel speech tech for small languages, using Fryslân as a living lab. (Lead: Igor Marchenko)

• Authentic human/machine interactions: We explore novel methods to recognize and synthesize non-literal speech, using sarcasm as case study. Techniques involve pioneering explorations into large language models like ChatGPT to synthesize data. Results will help people be understood by machines, even when they aren’t speaking literally. Some outputs: [1], [2], [3]; in the media: [1], [2], [3] (Leads: Xiyuan Gao and Zhu Li)

Leadership in the Speech Tech Community

Speech Tech Lab members serve in various roles and capacities in the community, indicative of our intention to foment social impact. Our members serve roles including:

• Organizers, Speech Synthesis Workshop / Blizzard Challenge 2025

• Ethics Committee, INTERSPEECH 2025

• Exhibition Committee, INTERSPEECH 2025

• Co-organizers, Dutch Speech Tech Day 2023 -- 2025

• Sponsor, Young Female Researchers in Speech INTERSPEECH workshop 2023 -- 2025

• Vice Chair, Language in the Human-Machine Era COST Action

• … among much more.

Work With Us

The Speech Technology Lab is always seeking opportunities to collaborate with researchers, industry partners, and community organizations. Whether you're interested in joint projects, student placements, or sponsored research, we welcome the chance to explore synergies that drive innovation in speech technology.

Research collaborations: Our interdisciplinary team has deep expertise spanning linguistics, machine learning, human-computer interaction, and more. We're eager to work with external partners on cutting-edge research projects that address real-world challenges in speech technology. If you have an idea for a collaborative study or grant proposal, we'd love to discuss it further.

Student opportunities We offer various opportunities for students to get involved with the work of the Speech Technology Lab, including:

• Research assistant positions for undergraduate and graduate students

• Thesis/dissertation supervision for students pursuing related topics

• Hosting postdoctoral researchers with external funding (e.g. MSCA staff exchange, COST scientific missions, and similar)

Please reach out to learn more about current openings and how to apply.

Sponsored Research The Speech Technology Lab is open to sponsored research agreements with industry, government, or non-profit organizations. If you're interested in funding targeted projects or initiatives within our scope of work, we'd be happy to discuss potential arrangements and deliverables.

Contact us

We welcome inquiries from potential collaborators, research partners, media, and anyone interested in the work we do at the Speech Technology Lab. Please feel free to contact our lab director, Dr Vass Verkhodanova.