A new simulation method for Time-Derivative Cellular Neural Networks

Time-derivative Cellular Neural Networks (TDCNN) is a general class of CNN having derivative connections between cells. It has been reported that several mixed-domain 3D spatiotemporal transfer functions for linear filtering can be implemented via TDCNN such as bandpass filters and visual receptive field models. In this paper a new computer simulation method for TDCNN is presented. The new simulation method reduces simulation time dramatically and simplifies the simulation steps.