As promised, we have published the updated method for luminance estimation and fitting function.
It is more robust for several reasons: mainly the function is now forced through zero and you can no longer get negative luminance values. We also have updated the loss function to deal with noise in a much more robust manner.
Let us know if you have any questions!
https://www.sae.org/publications/technical-papers/content/2022-01-0802/
Comments