ReadN'Karaoke

An adaptive aid that provides visual prosodic cues to assist young children when reading aloud

Lab Members: Heather Kember, Rupal Patel

Collaborators: Isabel Meirelles (Northeastern University), David Small (Small Design Firm), Sheelah Sweeny (Northeastern University)

Funding: Northeastern University Provost Award, NSF Grant No. HCC-0915527

Learning to read aloud in a fluent and expressive manner can be challenging. An important component of reading fluency is prosody, the melodic aspect of speech. Written text lacks sufficient cues as to the intended prosody and thus readers must infer these cues from context and punctuation. The ReadN'Karaoke software, developed in our lab, provides visual cues that correspond to multiple prosodic cues (F0, intensity, duration, and combinations of these features) to aid early readers in decoding the underlying prosodic intent.

The acoustic properties of an adult model reading the text serve as prosodic targets. These targets are rendered as continuously varying visual cues that correspond to the adult reader's prosodic variation aligned with the spoken text. These visual cues are presented in the context of an electronic storybook in which the reader can also record and play back their reading attempts. We are now assessing the usability of the software with a group of typically developing readers to examine the impact of visual prosodic cues on expressivity and comprehension.