Spiking network apparatus and method with bimodal spike-timing dependent plasticity
Apparatus and methods for learning in response to temporally-proximate features. In one implementation, an image processing apparatus utilizes bi-modal spike timing dependent plasticity in a spiking neuron network. Based on a response by the neuron to a frame of input, the bi-modal plasticity mechanism is used to depress synaptic connections delivering the present input frame and to potentiate synaptic connections delivering previous and/or subsequent frames of input. The depression of near-contemporaneous input prevents the creation of a positive feedback loop and provides a mechanism for network response normalization.
1. A method of updating a connection providing stimulus to an artificial spiking neuron, the method comprising:
determining a spike time based on a first spike of the neuron generated based at least in part on at least one stimulus received via at least one of a plurality of connections to the neuron;
determining a depression interval based at least in part on the spike time, the depression interval having a first boundary that is set at a pre-determined time before the spike time and a second boundary that is set at a pre-determined time after the spike time;
depressing a connection of the plurality of connections when a stimulus from the connection is received within the depression interval, the depression being configured to increase a time between a second spike and the first spike; and
potentiating a connection of the plurality of connections when a stimulus from the connection is received outside the depression interval, the potentiation being configured to decrease the time between the second spike and the first spike.
2. The method of claim 1 , wherein:
each connection of the plurality of connections comprise a latency parameter configured to provide a delay for transmissions of stimulus to the neuron;
potentiating the connection comprises reducing the latency parameter to decrease the delay; and
depressing the connection comprises increasing the latency parameter to increase the delay.