Difference of gaussian with temporal kernel

[jonathanoesterle]

Fitting a difference of gaussian with a temporal filter could be a nice feature for easy quantification of receptive fields.

[jonathanoesterle]

See Dayan and Abbot p 77:
D_s(x, y) = +- (1/(2 pi s_c²) exp(-(x²+y²) / 2 s_c²) - B/(2 pi s_s²) exp(-(x²+y²) / 2 s_s²) ) with s_s > s_c