Table 1.

Comparison of the test accuracy of our P²M enabled spiking CNN models with the baseline spiking CNN counterparts, where “MP” denotes membrane potential, “Custom conv.” denotes the incorporation of the non-ideal model to the ML algorithmic framework, and “Reduced dimensionality” denotes the reduction in the number of channels in the first convolutional layer.

Dataset	MP 1st layer	Custom conv.	Reduced dimensionality	Accuracy (%)
DVS128-Gesture	✓	×	×	93.40
DVS128-Gesture	×	×	×	88.78
DVS128-Gesture	×	✓	×	88.54
DVS128-Gesture	×	✓	✓	88.36
NMNIST	✓	×	×	98.10
NMNIST	×	×	×	93.68
NMNIST	×	✓	×	93.44
NMNIST	×	✓	✓	93.12