4 ms·
Difference with Gabors kernels (that are in our visual cortex and what the first level CNN kernels look to converge to) is the scaling factor 1/t instead of e^(
by trhway 1y ago
Difference with Gabors kernels (that are in our visual cortex and what the first level CNN kernels look to converge to) is the scaling factor 1/t instead of e^(-t)