Tracking Dynamics of Musical Engagement Using EEG
Self-reports of engagement or preference can suffer from response bias, and the act of delivering a continous behavioral response can distract from the stimulus at hand. We explore objective measures of musical engagement using inter-subject correlations of EEG and other physiological responses. We seek to show that the temporal organization of acoustical events into music is critical to achieving listener engagement, and to uncover relationships between musical engagement and repetition, anticipation, and expectation.
Dmochowski et al. (2014). Deriving the Neural Signatures of Musical Features Using Canonical Correlation Analysis. Spoken presentation at CogMIR: Cognitively Based Music Informatics Research, Toronto, Canada.
Kaneshiro et al. (2014). Toward an Objective Measure of Listener Engagement with Natural Music Using Inter-Subject EEG Correlation. In Proceedings of the 13th International Conference on Music Perception and Cognition and the 5th Triennial Conference of the Asia- Pacific Society for the Cognitive Sciences of Music, Seoul, Korea. [pdf]
Representational Similarity Analysis
Representational Similarity Analysis (RSA) applies scaling and clustering methods to pairwise distances of a set of stimuli, responses, or models to investigate the underlying structure of the set and compare representations across modalities. We use single-trial EEG classification to derive multi-category confusion matrices as input for RSA. This technique serves to reveal the dynamics of the representational space over the time course of the brain response, and also to identify spatial and temporal components of interest in the response.
Kaneshiro et al. (2015). A Representational Similarity Analysis of the Dynamics of Object Processing Using Single-Trial EEG Classification. PLoS ONE 10:8, e0135697. doi:10.1371/journal.pone.0135697 [web]
Kaneshiro et al. (2014). Visual Object Categories and Exemplars Can Be Decoded from Single-Trial EEG. Poster presentation at the Association for Psychological Science 26th Annual Convention, San Francisco, USA.
Kaneshiro et al. (2012). An Exploration of Tonal Expectation Using Single-Trial EEG Classification. In Proceedings of the 12th International Conference on Music Perception and Cognition and the 8th Triennial Conference of the European Society for the Cognitive Sciences of Music, Thessaloniki, Greece. [pdf]
Large-Scale Musical Engagement
As an alternative approach to studying musical engagement, I've been expanding the scope of my research to include larger datasets collected under controlled experimental conditions, as well as industrial-scale data. For the QBT-Extended dataset, we collected and published over 3,000 tapped representations of excerpts from popular songs. Projects in progress with Shazam have involved analysis of over 50M data points.
Kaneshiro et al. (2015). Large-Scale Music Discovery Behavior: Effects of Genre and Geography. Poster presentation at the biennial meeting of the Society for Music Perception and Cognition, Nashville, USA.
Kaneshiro et al. (2013). QBT-Extended: An Annotated Dataset of Melodically Contoured Tapped Queries. In Proceedings of the 14th International Society for Music Information Retrieval Conference, Curitiba, Brazil. [pdf] [web] [github]
Research Tools and Datasets
With the goal of promoting reproducible research and facilitating cross-disciplinary research, our group is moving in the direction of making our experimental data and analysis software publicly available to the research community. This includes collections of behavioral and EEG responses; mobile applications used to conduct research; and analysis and visualization tools used to arrive at published results.
Kaneshiro et al. (2015). Object Category EEG Dataset. Stanford Digital Repository. [web]
Kaneshiro et al. (2015). EEG-Recorded Responses to Short Chord Progressions. Stanford Digital Repository. [web]
Kaneshiro et al. (2015). EEG data analyzed in "A Representational Similarity Analysis of the Dynamics of Object Processing Using Single-Trial EEG Classification". Stanford Digital Repository. [web]
Kim et al. (2012). Tap-It: An iOS App for Sensori-Motor Synchronization (SMS) Experiments. In Proceedings of the 12th International Conference on Music Perception and Cognition and the 8th Triennial Conference of the European Society for the Cognitive Sciences of Music, Thessaloniki, Greece. [pdf] [github]