Sari Ylinen, associate professor in logopaedics, has an impressive, multidisciplinary background. She has studied languages and has PhD in psychology, she is a docent in cognitive neuroscience, she has conducted research on language education, and currently she works at Tampere University, the Faculty of Social Sciences, leading the Logopaedics team.
Sari’s research interests are focussed on language, brain, and learning, and the projects she is currently involved in reflect her multidisciplinary background. One of them is NordForsk-funded four-year project, TEFLON (Technology-enhanced foreign and second-language learning of Nordic languages). The project aims to help immigrant children, who need to learn one of the challenging Nordic languages, by developing a language-learning game targeting these languages (for previous work, click here).
An essential part of the game is automatic speech recognition, which can be used to assess children’s speech and provide feedback on it to reinforce learning. In addition to its main aim, this interdisciplinary project also aims to stimulate novel practices in engineering and language education. The project team consists of experts in engineering, cognitive neuroscience, psycholinguistics, speech-language pathology and language education.
The game is used also in another project that aims to provide a modified game version for children coming to Finland from Ukraine to facilitate their language learning and adaptation. In the project funded by Faculty of Social Sciences, Tampere University, Sari’s team has collected speech samples from children to attune automatic speech recognition for Ukrainian children by using deep neural networks.
Sari has also worked with newborn infants and small children, ages 1-2 years. By recording electroencephalography (EEG), she and her team have shown that infant brain can actually predict syllables that comprise words and that such predictive processing may assist in word recognition and learning at the early stages of language development. In fact, prediction and learning are closely intertwined in the brain.
- I am excited about our recent findings on infants’ learning. By using speech stimulation that enables infant brain to generate predictions, we can actually use EEG to measure newborn infants’ learning ability. This ability persists at later stages of development. Infants who can form better neural representations from speech exposure at birth produce longer utterances at 2 years of age. The link between early learning and later language ability suggests that in the future these methods may help to assess risk for language disorders or language delay soon after birth and enable early treatment and support.
In her new position, Sari is looking for new collaborators, who are interested in language development or language disorders as well as technology use in rehabilitation.
- Neurocenter Finland seems a very useful network for a newcomer like me. I already got some new contacts in Tampere through Tampere Brain & Mind, Sari says.
Tampere Brain & Mind is part of a nation-wide Neurocenter Finland researcher network. For more information, please visit www.neurocenterfinland.fi
Photographer: Anna-Riikka Smolander