Okko Räsänen

Interactive visualizations in TSExplorer allow researchers to annotate and explore time-series data more efficiently, transforming how we handle high-dimensional datasets.

Combining text and audio features reveals a significant boost in modeling emotional valence, but not arousal, in Finnish speech.

Self-supervised and visually grounded models are closing the gap in explaining how infants learn language from raw acoustic and visual input, challenging the need for strong linguistic priors.