Okko      Research     Publications      Resources      Contact
Okko Räsänen
      Okko Räsänen,
      D.Sc. (tech.), Title of Docent
      Research Fellow

      Department of Signal Processing and
      Acoustics, Aalto University, Finland
I'm a researcher at Aalto University, Department of Signal Processing and Acoustics, Finland, where I work with our Language Acquisition Group (LAG).

My main research interests lie at the interaction of human and machine cognition, and therefore I consider myself as a cognitive scientist with a background in language technology and cognitive psychology. I primarily use computational models (with signal processing and machine learning) to understand human information processing, and, on the other hand, I try apply findings from cognitive science of spoken language to the development of intelligent machines and algorithms that could deal with the complexity of the external world.

My primary topical interests lie in language processing. I'm especially interested in how human children can (or how machines could) learn meaningful language representations from the sensorimotor experiences that they can acquire during the early development, and how this learning impacts their further language learning, perceptual attention, and overall behavior. This also includes questions such as how is language connected to the sensorimotor world beyond the structure and symbols of language itself, and how do these connections shape the way how languages are learned in childhood.

However, my research interests also cover a broad range of other topics related to spoken language processing in humans and machines, including topics such as paralinguistic speech processing (~speech contents beyond the literal spoken message), zero-resource speech processing, psychoacoustics, speaking style conversion, and modeling and analysis of suprasegmental structure of acoustic speech.

May 2018

Why infants listen more to infant-directed (IDS) than adult-directed speech (ADS)? Is prosodic learning from IDS more efficient than from ADS? We tried to address these questions in our recent paper accepted for publication in Cognition, available at https://psyarxiv.com/um6d7/.

Full citation:

Räsänen, O., Kakouros, S. & Soderstrom, M. (in press). Is infant-directed speech interesting because it is surprising? — Linking properties of IDS to statistical learning and attention at the prosodic level. Cognition, accepted for publication (.pdf).

Jan 2018

Our study on the role of statistical learning in the perception of prominence in speech has now been published in Neuropsychologia. In that paper, we show how F0 cues for prominent words can be reversed by exposing listeners to stimuli with certain statistical structure at the prosodic level. Check it out at https://psyarxiv.com/7ywzd/.

Full citation:

Kakouros, S., Salminen, N. & Räsänen, O. (2018). Making predictable unpredictable with style - Behavioral and electrophysiological evidence for the critical role of prosodic expectations in the perception of prominence in speech. Neuropsychologia, 109, 181–199

October 2017

Our new journal article on how infants might perceive syllable-like units before they learn their native language was accepted for publication in Cognition. The pre-prints and Matlab scripts for syllabification are available at https://osf.io/86wmj/.

Full citation:

Räsänen, O., Doyle, G., & Frank, M. C. (2018). Pre-linguistic segmentation of speech into syllable-like units. Cognition, 171, 130–150

Selected publications

Kakouros, S. & Räsänen, O. (2015). Perception of sentence stress in speech correlates with the temporal unpredictability of prosodic features. Cognitive Science, 40, 1739–1774 (web).

Räsänen, O. & Rasilo, H. (2015). A joint model of word segmentation and meaning acquisition through cross-situational learning. Psychological Review, 122(4), 792–829 (.pdf).

Räsänen, O. & Laine, U. K. (2013). Time-frequency integration characteristics of hearing are optimized for perception of speech-like acoustic patterns. The Journal of the Acoustical Society of America, 134, 407–419 (web).

Räsänen, O. (2011). A computational model of word segmentation from continuous speech using transitional probabilities of atomic acoustic events. Cognition, 120, 149–176 (web) (.pdf).

Contact: firstname.surname@aalto.fi