Peter Härtl: 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
91.7%z 5.15 - 2
89.3%z 4.78 - 3
81%z 4.10 - 4
66.6%z 3.87 - 5
65.9%z 3.86 - 6
47.4%z 3.62 - 7
47.2%z 3.62 - 8
46.4%z 3.56 - 9
45.2%z 3.49 - 10
44.5%z 3.44 - 11
43.3%z 3.36 - 12
42.7%z 3.32 - 13
42.5%z 3.31 - 14
42.3%z 3.30 - 15
42.2%z 3.29 - 16
41.2%z 3.23 - 17
41%z 3.21 - 18
40.8%z 3.20 - 19
40.7%z 3.20 - 20
40.6%z 3.19 - 21
40.6%z 3.19 - 22
40.4%z 3.17 - 23
39.9%z 3.14 - 24
39.6%z 3.13 - 25
39.6%z 3.12 - 26
39%z 3.08 - 27
39%z 3.08 - 28
38.9%z 3.08 - 29
38.9%z 3.08 - 30
38.7%z 3.07 - 31
38.4%z 3.04 - 32
38.4%z 3.04 - 33
38.4%z 3.04 - 34
38%z 3.02 - 35
37.9%z 3.01 - 36
37.9%z 3.01 - 37
37.9%z 3.01 - 38
37.2%z 2.97 - 39
37%z 2.96 - 40
36.8%z 2.94 - 41
36.7%z 2.93 - 42
36.7%z 2.93 - 43
36.5%z 2.92 - 44
36.5%z 2.92 - 45
36.3%z 2.91 - 46
36.3%z 2.91 - 47
35.7%z 2.87 - 48
35.6%z 2.87 - 49
35.5%z 2.85 - 50
35.4%z 2.85