Fig. 4: BCM learning rule for pattern learning based on device-level artificial striate cortex.
From: Synaptic plasticity in self-powered artificial striate cortex for binocular orientation selectivity

a The schematic illustration of artificial striate cortex and the postsynaptic electric circuit design. b The optical images of hemispherical-shaped artificial striate cortex and the flexible array glowing as sunlight passes through it (insert). And the dimension of the array is 3 cm × 3 cm with the electrode width of 1000 μm, where the large array size is limited by the size of sub-solar cell for receiving sufficient visual information from the environment. c The schematic of circuit block. d The measured (triangle) and fitting (line) results of BCM learning rule as a function of postsynaptic firing frequency based on the artificial striate cortex. e The schematic of artificial striate cortex for pattern learning. f The evolution of synaptic weights of the 9 visual cortical synapses during pattern learning task. g The initial and final states of synaptic weights in the pattern learning task.