Language relies on perception for accurate identification of both oral and written speech and memory for binding of these arbitrary sensory stimuli to semantic content. This relationship has social, cultural and biomedical implications at a variety of scales from the individual to the global. For example, early exposure to the sounds of a second language is often correlated with developing fluency in that language, a problem with significant implications for interactions between individuals in different cultures. Deficits in both visual and auditory processing contribute to language impairments (e.g., dyslexia, hearing loss), with attendant implications for communication and education worldwide. Not only does perception play a causal role in language comprehension, but there may also be an arrow of causation running the other way. Recent theories of language function in the brain have suggested that words may evoke a kind of mental simulation of a remembered perceptual experience connected to the meaning of that word. For example, words like “kick” have been shown to activate the foot region of the motor cortex. In short, language may work by co-opting perceptual and action networks in the brain. This project team examined these concepts using fMRI imaging and psychophysical techniques.

A study explored a novel metric for evaluating piano performance. Quantifying musical performance is time-consuming, especially because there are few objective, automatable methods to implement this analysis. Dynamic time warping (DTW) is an algorithm for measuring similarity between two temporal sequences that can vary in length or speed. Because DTW can determine the similarity between two temporal sequences regardless of length, it is often used to analyze motor movements with a stereotyped pattern. Applying DTW to trace the acquisition of a novel piano piece, team members found a strong correlation between DTW score and number of errors qualitatively determined, suggesting that DTW is a reliable metric for quantifying proficiency in piano performance. It is also time-efficient – more than 1,000 pieces can be scored through DTW in the average time taken to score a single piece (48 seconds long) qualitatively.

Timing

Summer 2014 – Spring 2015

Team Outcomes

Dynamic Time Warping: A Novel Metric for Evaluating Piano Performance (poster by Manish Nair and Anya K. Ranganathan, with Jonathan Adler, Wilson Brace, Jennifer Groh, Tobias Overath, Zab Johnson)

This Team in the News

Duke Seniors Share What Was Most Meaningful about Their Bass Connections Experiences

See earlier related team, Perception, Language and Memory (2013-2014).