Here are brief descriptions of my research projects in reverse chronological order.
-
Spring 2008 - Present,
Melody Familiarity and Recognition Network, Dissertation Proposal Pilot Study
Advisor: Dr. Anthony Maida
Designed and implemented a Tank and Hopfield-based musical sequence
detection neural network in Java. Currently using the network to illustrate the effects of musical training
on familiarity judgments of melodies, by representing long term memory for musicians
and nonmusicians using separate melody detection networks and comparing performance.
Initial results from this computational model were presented at the Neurosciences and Music-III conference in Montreal, June 26th-28th, 2008.Abstract. Poster.
-
Spring 2007,
RoboSim, Ph.D. II year Course Project
Implemented a Monte Carlo Localization algorithm, as part of a 2-member team with
Derek James, in Java, for updating a
robot’s belief about its state. Painted test bed, obstacles, robot and sensors using Swing and Java 2D API.
-
Fall 2006,
Modeling Musical Scale Learning and Perception using Neural Networks, Ph.D. II year Course Project
Advisor: Dr. Anthony Maida
Implemented a musical scale-classification feed-forward neural network, in MatLab to
illustrate aspects of music cognition pertaining to key invariance and musical
training.
An Elman feed-forward network with context units was used for the purpose of this project in order to
capture a melody temporally.
-
Spring 2006,
Musician-influenced Bisociation in the Listener, Ph.D. I year Course Paper/Project
Used specific examples in music, to extend Dasgupta's (2006) "doctrine of the creative encounter"
by hypothesizing that one of the ways in which a musician brings about change in the listener's
cognitive identity is by forcing unintentional bisociation in the listener through his or her act
of creation, with or without the knowledge of the listener. Furthermore, the type of bisociation experienced
by the listener is largely similar to that experienced by the musician prior to and during his or her act of creation.
-
Spring 2006,
Testing a Model of Perceptual Specialization for Faces, Ph.D. I year Project
Supervisor: Dr. Mike Kalish
Implemented a Mixture of Experts feed-forward classification network for understanding issues
in cognitive science pertaining to face perception and object perception, in MatLab.
The project involved: a) Reviewing existing computational models in face/object
perception, b) Selecting an existing computational model, c) Implementing an
interpretation of the existing model, followed by alterations/modifications.
-
Fall & Summer 2005, A Computational Model of the Music of Stevie Ray Vaughan,
Master's thesis
Advisor: Dr.
Subrata Dasgupta
Having spent several years of my professional life in
non-music related activity, I finally felt the irresistible urge to integrate my
interest in blues and blues-rock music with computer science. My master's thesis
involved the design of a probabilistic computational model of Stevie Ray
Vaughan's music. The model generated predictions of Stevie Ray Vaughan licks from user-based input.
An abstract of my thesis is available
here. I presented this work as a full paper at the 6th Creativity and
Cognition conference in DC, June 13th-15th, 2007. You may
download a copy of this paper here. I also
presented my thesis on two occasions at the University of Louisiana, Lafayette.
Contact me for a copy of this
talk.