emiwar / caJulia

Fast CNMF-E implementation with Julia and CuArrays
0 stars 0 forks source link

Residual image #3

Open emiwar opened 2 years ago

emiwar commented 2 years ago

Pixel-wise mean, max, min and negentropy of residual likely has interesting info. If neuron-size bump in min then we should split footprint (I think).

emiwar commented 2 years ago

Maybe mean should always be 0 because it's absorbed in the static background?