Number of simulation markers

• by mmalec

  1 in 10 of our images are simulations. Are the simulations evenly split between fake lenses and images know to not have lenses? I'm trying to get a feel for how many of my potential lenses count is due to marked simulations.

  Posted June 13, 2013 9:01 PM

• by Capella05 moderator

  Not necessarily - but a good rule of the thumb is 10%

  Posted June 15, 2013 2:40 PM

• by anupreeta scientist, admin

  The training images consists of sim lenses and dud images (ie. images checked to have no known lenses). They roughly have comparable proportions and they are roughly shown with similar frequencies.

  Posted June 19, 2013 3:05 PM