Skip to main content
. 2021 Jul 27;21(15):5068. doi: 10.3390/s21155068
Algorithm 1 Evaluating models to differentiate siblings.
Input: Image Pairs of Siblings M with m pairs
τ = threshold array defined for each similarity measure
mod = respective model
for i = 1 to m do
      for each image in pair do
         I1 → read image;
         I1 → resize image;
         I1 → pre-process image;
         P1 = extractEmbeddings(mod(I1));
         I2 → read image;
         I2 → resize image;
         I2 → pre-process image;
         P2 = extractEmbeddings(mod(I2))
      end
end
cosine cosine_distance(P1, P2)
euc → euclidean_distance( P1, P2 )
ssim → structured_similarity( P1, P2 )
manh manhattan_distance( P1, P2 )
mink → minkowski_Distance( P1, P2 )
for j = 1 to τ.length do
      if distance < τ[j]
         distance_result → “Same”
      else
         distance_result → “Different”
end