Project description: general tips
I will be continuing the work that I have been doing for the past year on the project entitled "Using neuroscience to categorize music." The project investigates real, objective measures that can be used to accurately predict human neurological responses to musical patterns. This would apply to a diverse amount of stimuli - as found in recordings from all types of music and as perceived by human listeners, from what a chef hears while preparing a meal from clinking of silverware and china to what a conductor hears while conducting an orchestral score from memory.
The goal of the project is to engineer a search engine for music based on real, perceptually salient, and quantitative aspects of music audio signals. The project combines research in audio and acoustic engineering, perception and cognition, computer science, and music. In simpler terms, it is taking a new way of analyzing musical rhythm patterns by separate segments of audio into independent components and organizing them into latent timbre channels.
The first step is encoding streams of feature data in these channels, thus encoding separate information about each aspect of the musical mixture. I will assist in the design, running, and analysis of behavioral studies in human rhythm perception - specifically the stimulus preparation, collecting and analyzing fMRI data, processing similarity matrices for computational analysis, and helping to design both computational algorithms that predict the behavioral responses to rhythmic aspects of music audio and also an interface for searching / browsing music by rhythmic content.
- Too long
- Multiple paragraphs instead of just one
- Includes the project title (don't do this!)
- Extraneous information (e.g. the chef/silverware and orchestra conductor sentence)
- Information that should be in other sections of the application
- the first sentence should be in "related experience"
- the last paragraph should be in "student role"
The goal of the project is to engineer a search engine for music based on real, perceptually salient, and quantitative components of music audio signals. To determine what components are salient to listeners, neurological responses to specific auditory signals will be measured by presenting auditory stimuli to research participants while they undergo functional magnetic resonance imaging (fMRI). The auditory stimuli will be created with known, categories, and organized rhythmic patterns and will be presented in different channels. Therefore, information about each aspect of the musical mixture can be encoded separately and the participants’ responses to specific stimuli can be analyzed and compared. Based on these computational analyses, algorithms will be developed that can predict behavioral responses to rhythmic aspects of music. The algorithms will then enable searching and browsing of music by rhythmic content.