Purin Mechanism Introduces Synaptic Efficacy to CNNs

Purin adds biology-inspired synaptic efficacy modulation to convolutional neural networks using a time-interval-based abstraction. It utilizes two weight matrices to represent input and output efficacy, allowing for short- and long-term changes without discrete time-step simulations.