Sparse Coding Using the Locally Competitive Algorithm on the TrueNorth Neurosynaptic System (bibtex)

by K.L. Fair, D.R. Mendat, A.G. Andreou, C.J. Rozell, J. Romberg and D.V. Anderson

Abstract:

The Locally Competitive Algorithm (LCA) is a biologically plausible computational architecture for sparse coding, where a signal is represented as a linear combination of elements from an over-complete dictionary. In this paper we map the LCA algorithm on the brain-inspired, IBM TrueNorth Neurosynaptic System. We discuss data structures and representation as well as the architecture of functional processing units that perform non-linear threshold, vector-matrix multiplication. We also present the design of the micro-architectural units that facilitate the implementation of dynamical based iterative algorithms. Experimental results with the LCA algorithm using the limited precision, fixed-point arithmetic on TrueNorth compare favorably with results using floating-point computations on a general purpose computer. The scaling of the LCA algorithm within the constraints of the TrueNorth is also discussed.

View PDF

Reference:

Sparse Coding Using the Locally Competitive Algorithm on the TrueNorth Neurosynaptic SystemK.L. Fair, D.R. Mendat, A.G. Andreou, C.J. Rozell, J. Romberg and D.V. Anderson. Frontiers in Neuroscience, vol. 13, pp. 754, July 2019.

Bibtex Entry:

@article{fair.19,
  author = 	 {Fair, K.L. and Mendat, D.R. and Andreou, A.G. and Rozell, C.J. and Romberg, J. and Anderson, D.V.},
  title = 	 {Sparse Coding Using the {L}ocally {C}ompetitive {A}lgorithm on the {TrueNorth} Neurosynaptic System},
  year =	 2019,
  month = jul,
  volume={13},      
  pages={754},     
  abstract = {The Locally Competitive Algorithm (LCA) is a biologically plausible computational architecture for sparse coding, where a signal is represented as a linear combination of elements from an over-complete dictionary. In this paper we map the LCA algorithm on the brain-inspired, IBM TrueNorth Neurosynaptic System. We discuss data structures and representation as well as the architecture of functional processing units that perform non-linear threshold, vector-matrix multiplication. We also present the design of the micro-architectural units that facilitate the implementation of dynamical based iterative algorithms. Experimental results with the LCA algorithm using the limited precision, fixed-point arithmetic on TrueNorth compare favorably with results using floating-point computations on a general purpose computer. The scaling of the LCA algorithm within the constraints of the TrueNorth is also discussed.},
  journal = {Frontiers in Neuroscience},
  url = {http://doi.org/10.3389/fnins.2019.00754}
}