Hakan Peker: raw computed results
The raw ranking from the production similarity calculation (face-embedding centroid cosine → gender-cohort z-normalization). The web-opinion crown, curation, and fusion re-rank are removed.
- 1
83.9%z 4.33 - 2
81.9%z 4.17 - 3
81.5%z 4.14 - 4
80.9%z 4.09 - 5
75.5%z 3.97 - 6
73.5%z 3.95 - 7
59.3%z 3.79 - 8
54.9%z 3.74 - 9
53.8%z 3.73 - 10
47.8%z 3.66 - 11
47.5%z 3.64 - 12
47.2%z 3.62 - 13
47%z 3.60 - 14
46.9%z 3.60 - 15
46.6%z 3.58 - 16
46.5%z 3.57 - 17
46.2%z 3.55 - 18
45.5%z 3.51 - 19
44.2%z 3.42 - 20
44%z 3.41 - 21
44%z 3.41 - 22
43.9%z 3.40 - 23
43.9%z 3.40 - 24
43.1%z 3.35 - 25
42.9%z 3.34 - 26
42.9%z 3.33 - 27
42.7%z 3.32 - 28
42.6%z 3.32 - 29
41.9%z 3.27 - 30
41.7%z 3.26 - 31
41.6%z 3.25 - 32
41.6%z 3.25 - 33
41.5%z 3.24 - 34
41.4%z 3.24 - 35
41%z 3.21 - 36
40.9%z 3.21 - 37
40.9%z 3.21 - 38
40.7%z 3.19 - 39
40.7%z 3.19 - 40
40.6%z 3.19 - 41
40.5%z 3.18 - 42
40.5%z 3.18 - 43
40.4%z 3.18 - 44
40.4%z 3.17 - 45
40.3%z 3.17 - 46
40.1%z 3.16 - 47
40.1%z 3.15 - 48
40%z 3.15 - 49
40%z 3.15