I'm interested in understanding the form and function of perceptual memory: how can prior perceptions be effectively remembered, recalled and used? My work has touched on the areas of machine learning, signal processing, and psychophysics, particularly as they pertain to the perception of sound. I use the methodology and experimental evidence from these fields to better understand human and machine learning of perceptual data.
I strive to incorporate both empirical and normative issues of perception in my work. How do people learn to perceive (empirical), and how should perception occur. The brain's function can utilize only limited resources and time. What can normative computer models teach us about the compromises nature has made? Furthermore it is evident that we do not yet understand how to build machines that perceive as effectively, or as comprehensively as humans do. What flawed assumptions can human behavior reveal about our computational models, and what computational strategies does human behavior suggest?
Below you can find the list of current and past projects I have worked on.
- Current Projects
- Past Projects
Current Projects
Computational Models of Auditory Perceptual Learning
With sufficient practice, human beings are able to enhance the acuity of their sensory systems. This is known in the literature as perceptual learning. Recent work in perceptual learning (e.g. Banai et al., 2009; Seitz et al., 2005; Yotsumoto et al., 2008; Zhang et al., 2008) , has shown that learning on one task (called the target) may be prevented when a second task (called the distractor) is practiced either during or shortly after practice of the target: this is called learning interference. These results suggest distinct properties of short and long term stores of perceptual memory because what interferes with learning during practice is distinct from what interferes after practice.
In this research I have been developing a computational model of several of these experiments (Banai et al., 2009; Yotsumoto et al., 2008) in an effort to better understand the form and function of short and long term stores of human perceptual memory. The longer term goal of this project is to develop an account of short and long term stores of memory consistent across a great variety of perceptual learning experiments. With the exception of the recent work in Sotiropoulos et al. (2011) little has been done to consider the ramifications of computational models of perceptual learning across more than just a few experiments.
Related Papers
- Little, D., Pardo, B., & Wright, B. A. (2011). A Computational Model of Auditory Perceptual Learning: Predicting Learning Interference Across Multiple Tasks. Proceedings of the 33rd Annual Conference of the Cognitive Science Society. Boston, MA.
- D. Little, B. Pardo. Computational Models of Perceptual Learning Across Multiple Auditory Tasks: Modeling Daily Learning Limits as Memory Decay, in Proceedings of the 10th International Conference on Cognitive Modeling, Philadelphia, PA, USA, August 5-8, 2010.
Past Projects
Learning sounds from mixtures of audio: 2007-2009
More information about this project can also be found at my lab's website.
My goal in this project was to understand how machine systems can learn to recognize sounds from mixtures of sound. More specifically, I wanted to design a machine system capable of learning to label sounds from mixtures of sounds, using only weak labels. A weak label is much like the label an infant might get: all the learner knows is that something, somewhere around a particular time is important, and the learner must figure out how to appropriately assign that label to future sounds. There may be multiple sounds during learning, which may overlap (in time) with the sound the label is associated with (which I call the target).
You can find examples of the sounds one version of my system was evaluated with, as described in my ISMIR, 2008 paper, here
Related Papers
- D. Little, B. Pardo, Learning musical instruments from mixtures of audio with weak labels. 9th International Conference on Music Information Retrieval (ISMIR) September 14th-18th, 2008, Philadelphia, USA.
Online learning for Query-by-Humming: 2006-2007
During my first year as a graduate student I worked on a project called VocalSearch, now named Tunebot. VocalSearch allowed a user to search for music by singing melodies into a microphone (also called Query-by-Humming). I showed that the system can improve results for particular users, after it has been deployed, based on feedback from the user. This feedback customizes the parameters of how notes are identified and compared in a person's sung melody.
Related Papers
- D. Little, D. Raffensperger, B. Pardo, A Query By Humming System that Learns from Experience, 8th International Conference on Music Information Retrieval (ISMIR) September 23-27, 2007, Vienna, Austria. (PDF)
- D. Little, D. Raffensperger, B. Pardo, User-specific Training for a Music Search Engine, 4th Joint Workshop on Multimodal Interaction and Related Machine Learning Algorithms, June 28-30, 2007, Brno, Czech Republic. (PDF)
- D. Little, D. Raffensperger, and B. Pardo. Machine Learning and Mutlimodal Interaction: Fourth International Workshop, MLMI 2007, Brno, CZ, June 28-30, 2007, Revised Selected Papers, chapter User specific training of a music search engine. Lecture Notes in Computer Science. Springer, 2007.
Hexagonal Metamorphic Robot Algorithms: 2003-2005
As an undergraduate, I worked for three years with Jennifer Walter at Vassar College developing algorithms for the rearrangement of hexagonal robots. The purpose of this research is to design a set of reconfigurable modular hexagonal robots, which would enable an adaptive system able to rearrange itself for various tasks in hostile environments. The work I did was largely theoretical: I looked at how to move a large collection of these robots efficiently and without error around various kinds of obstacles.
Related Papers
- D. Little and J. Walter, Using Hexagonal Metamorphic Robots to Form Temporary Bridges, in Proc. of the IEEE International Conference on Intelligent Robotic Systems , Aug. 2005, Edmonton, Alberta, Canada, pages 2652-2657.
- J. Walter, M. Brooks, D. Little, and N. Amato, Enveloping Multi-Pocket Obstacles with Hexagonal Metamorphic Robots, in Proc. of the IEEE Intl. Conf. on Robotics and Automation, Apr. 2004, New Orleans, LA, pages 2204-2209.
- J. Walter and D. Little, Bridging Gaps in Traversal Surfaces with Hexagonal Metamorphic Robots, in Proc. of the American Nuclear Society 10th International Conference on Robotics and Remote Systems for Hazardous Environments, 28-31 March, 2004, Gainsville, FL.
References
- K Banai, J A Ortiz, J D Oppenheimer, and B A Wright. Learning two things at once: differential constraints on the acquisition and consolidation of perceptual learning. Neuroscience, 2009.
- A R Seitz, N Yamagishi, B Werner, N Goda, M Kawato, and T Watanabe . Task-specific disruption of perceptual learning. Proceedings of the National Academy of Sciences, 102 (41): 14895-14900, 2005. doi: rm10.1073/pnas.0505765102
- Grigorios Sotiropoulos, Aaron R Seitz, and Peggy Series. Perceptual learning in visual hyperacuity: a reweighting model. Vision Research, In Press, 2011.
- Yuko Yotsumoto, Takeo Watanabe, and Yuka Sasaki. Different dynamics of performance and brain activation in the time course of perceptual learning. Neuron, 57 (6): 827-33, March 2008. ISSN 1097-4199. doi: rm10.1016/j.neuron.2008.02.034.
- J Y Zhang, S G Kuai, L Q Xiao, S A Klein, D M Levi, and C Yu. Stimulus coding rules for perceptual learning. PLoS Biol, 6 (8): e197, 2008. doi: rm10.1371/journal.pbio.0060197.