A Hardware/Software Framework for Real-Time Spiking Systems
One focus of recent research in the field of biologically plausible neural networks is the investigation of higher-level functions such as learning, development and modulatory functions in spiking neural networks. It is desirable to explore these functions in physical neural network systems operating in real-time. We present a framework which supports such research by combining hardware spiking neurons implemented in analog VLSI (aVLSI) together with software agents. These agents are embedded in the spiking communication of the network and can change the parameters and connectivity of the network. This new approach incorporating feedback from active software agents to aVLSI hardware allows the exploration of a large variety of dynamic real-time spiking network models by adding the flexibility of software to the real-time performance of hardware.
Unable to display preview. Download preview PDF.
- 2.Deiss, S.R., Douglas, R.J., Whatley, A.M.: A pulse-coded communications infrastructure for neuromorphic systems. In: Maass, W., Bishop, C.M. (eds.) Pulsed Neural Networks, pp. 157–178. MIT Press, Cambridge (1999)Google Scholar
- 3.Vogelstein, R., Mallik, U., Cauwenberghs, G.: Beyond address-event communication: dynamically-reconfigurable spiking neural systems. In: The Neuromorphic Engineer, vol. 1. Institute of Neuromorphic Engineering, INE (2004)Google Scholar
- 4.Douglas, R., Mahowald, M.: Silicon neurons. In: Arbib, M. (ed.) The Handbook of Brain Theory and Neural Networks, pp. 282–289. MIT Press, Boston (1995)Google Scholar
- 6.Dante, V., Del Giudice, P.: The PCI-AER interface board. In: Cohen, A., Douglas, R., Horiuchi, T., Indiveri, G., Koch, C., Sejnowski, T., Shamma, S. (eds.) 2001 Telluride Workshop on Neuromorphic Engineering Report, pp. 99–103 (2001)Google Scholar
- 8.Oster, M., Liu, S.C.: A winner-take-all spiking network with spiking inputs. In: 11th IEEE International Conference on Electronics, Circuits and Systems, ICECS (2004)Google Scholar
- 9.IST-2001-34124: Caviar - convolution aer vision architecture for real-time (2002), http://www.imse.cnm.es/~bernabe/CAVIAR