Beat gestures

Overview

This project investigated whether visual beat gestures can help listeners segment continuous speech into words when acoustic cues are ambiguous or unreliable.

In spoken language comprehension, listeners often need to interpret a continuous speech stream without clear boundaries between words. When acoustic information is reduced or difficult to rely on, they may need to use other available cues to resolve uncertainty. Beat gestures provide a potential visual cue: their rhythmic timing may highlight certain parts of the speech stream and help listeners decide where word boundaries are likely to occur. In this sense, I found it closely connected to my broader interest in ambiguity resolution.

Reflection

This project was my second experience of going through a complete cycle of research training, from experimental design and data collection to statistical analysis and report writing. It was therefore not only a valuable project in multimodal speech perception, but also an important step in my development as a researcher.

What I found most meaningful was that the project helped me see connections across seemingly different research topics: identifying a theoretical question, translating it into an experimental design, testing it with data, and interpreting the results carefully. This experience made me realise that, across different domains, research often shares a common structure of thinking.

Because of this, the project gave me a stronger sense of confidence. It made me feel that I have the ability to investigate any question that genuinely interests me, even when the topic is initially unfamiliar. In this sense, the project strengthened both my research skills and my trust in my own curiosity.

Project period

December 2025 – May 2026

Supervision

This project was conducted under the supervision of Dr. Hans Rutger Bosker.

Read more

Find more interesting demos through the lab website: Speech Perception in Audiovisual Communication (SPEAC)!