Columnar-Organized Memory (COM): Brain-inspired associative memory with large capacity and robust retrieval

Abstract

Abstract Brain-inspired computing is an interdisciplinary field, which aims to use the capabilities of the human brain. The brain is remarkable for large memory capacity and robust memory retrieval. These characteristics are due to the specific structure of the brain. In this paper, an associative memory with large capacity and robust retrieval is introduced that is inspired by the cortical structure. The proposed model is called Columnar-Organized Memory (COM). Its architecture is based on the spiking winner-take-all (WTA): There are lateral excitatory connections between the WTAs and lateral inhibitory connections between the neurons of a WTA. Training of COM involves a two-phase algorithm: pattern storage phase and pattern association phase. The pattern storage phase is based on a spike timing dependent plasticity (STDP) rule and the pattern association phase is compatible with the classic Hebbian rule. The key characteristics of the proposed model are storing a large number of messages and retrieving them accurately. These properties are related to its structure notably; robust retrieval is a consequence of the lateral excitatory connections. Also, the overall effect of the lateral inhibitory connections is large storage capacity.