CSRNet-pytorch
CSRNet-pytorch copied to clipboard
Density Map generation
In density map generation, the gaussian_filter_density function will lead to slight variation. So i change the code:
density += scipy.ndimage.filters.gaussian_filter(pt2d, sigma, mode='constant')
to
map = scipy.ndimage.filters.gaussian_filter(pt2d, sigma, mode='constant') map = map / map.sum() density += map
This could avoid the slight variation, but would it affect the training?
the map.sum() is 1,I think it didn't work, did it?