This is the code for a neural network model of how infants learn to produce canonical babbling.

The code is written in MATLAB. It requires Praat (http://www.fon.hum.uva.nl/praat/) to be installed and was developed for Mac and Windows.

The model is closely related to those described in these papers:

A. S. Warlaumont, “Salience-based reinforcement of a spiking neural network leads to increased syllable production,” in IEEE International Conference on Development and Learning and Epigenetic Robotics (ICDL), 2013.

A. S. Warlaumont, “A spiking neural network model of canonical babbling development,” in Proceedings of the 2012 IEEE International Conference on Development and Learning and Epigenetic Robotics (ICDL), 2012.

It borrows heavily from Izhikevich's (2007) model:

E. M. Izhikevich, “Solving the distal reward problem through linkage of STDP and dopamine signaling,” Cerebral Cortex, vol. 17, no. 10, pp. 2443–2452, 2007. Code available at http://izhikevich.org/publications/dastdp.htm

And from Coath et al.'s model of auditory salience:

M. Coath, S. L. Denham, L. Smith, H. Honing, A. Hazan, P. Holonwicz, and H. Purwins, “An auditory model for the detection of perceptual onsets and beat tracking in singing,” in Neural Information Processing Systems, Workshop on Music Processing in the Brain, 2007. Code available at http://emcap.iua.upf.edu/downloads/content_final/auditory_saliency_model.html

Authors: Anne S. Warlaumont and Megan K. Finnegan