December 2014
despite the title there are no more references to python in this talk.
write documentation with embedded analysis
vs
This talk is about something else hence the title but…
Suposedly a set of
Image quality metrics from PreFace by Aware Software
Query.Name | Query.ID | Target.Name | Target.ID | Score |
---|---|---|---|---|
A0000_0000_00_C_G_Q.jpg | 1 | A0000_0000_00_C_G_Q.jpg | 1 | 1.000000 |
A0000_0000_00_C_G_Q.jpg | 1 | A0005_0000_00_C_G_Q.jpg | 6 | 0.996440 |
A0000_0000_00_C_G_Q.jpg | 1 | A0199_0000_02_C_G_Q.jpg | 200 | 0.567633 |
A0000_0000_00_C_G_Q.jpg | 1 | A0118_0000_02_C_G_Q.jpg | 119 | 0.560839 |
A0000_0000_00_C_G_Q.jpg | 1 | A0138_0000_02_C_G_Q.jpg | 139 | 0.560240 |
A0000_0000_00_C_G_Q.jpg | 1 | A0665_0000_05_C_G_Q.jpg | 666 | 0.551213 |
A number of images failed quality control, face could not be detected which left
extracted subset with common target and query filenames convert to a matric
\[ \left( \begin{array}{ccc} s_{1,1} & s_{1,2} & s_{1,3} & ... & s_{1,i} & ... & s_{1,n} \\ s_{2,1} & s_{2,2} & s_{2,3} & ... & s_{2,i} & ... & s_{2,n} \\ s_{3,1} & s_{3,2} & s_{2,3} & ... & s_{3,i} & ... & s_{3,n} \\ ... & ... & ... & ... & ... & ... & ... \\ s_{i,1} & s_{i,2} & s_{i,3} & ... & s_{i,i} & ... & s_{i,n} \\ ... & ... & ... & ... & ... & ... & ... \\ s_{n,1} & s_{n,2} & s_{n,3} & ... & s_{n,i} & ... & s_{n,n} \end{array} \right) \]
where \(s_{i,j}\) is the similarity score of image \(i\) as query with image \(j\) as target
Select images where
and in doing this it results in data where
i.e. the matrix is symmetric
A set of 3155 Images
If first we convert the 0-1 similarity score
to
a 0-1 distance score with
\(d_{i,j}=1-s_{i,j}\)
then cluster the results with hclust
the resulting dendrogram the resulting subtree
Removing images with no age gives
A set of 1726 Images
Linear Model \(similarity \thicksim age * \delta age * sex\) has an adjusted \(R^{2}\) of 0.32 for subject self comparisons scores
Missing child:
0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | |
---|---|---|---|---|---|---|---|---|---|
0 | 14:28 | 0:0 | 0:0 | 0:0 | 0:0 | 0:0 | 0:0 | 0:0 | 0:0 |
1 | 0:3 | 0:3 | 0:0 | 0:0 | 0:0 | 0:0 | 0:0 | 0:0 | 0:0 |
3 | 0:1 | 0:1 | 0:1 | 0:1 | 0:0 | 0:0 | 0:0 | 0:0 | 0:0 |
4 | 2:2 | 2:2 | 1:1 | 1:1 | 1:1 | 0:0 | 0:0 | 0:0 | 0:0 |
5 | 64:88 | 7:23 | 7:23 | 8:23 | 8:23 | 9:23 | 0:0 | 0:0 | 0:0 |
6 | 131:146 | 88:106 | 11:30 | 11:30 | 11:29 | 14:29 | 13:29 | 0:0 | 0:0 |
7 | 1:1 | 1:1 | 1:1 | 1:1 | 1:1 | 0:0 | 0:0 | 0:0 | 0:0 |
8 | 1:1 | 1:1 | 1:1 | 1:1 | 1:1 | 1:1 | 0:0 | 0:0 | 0:0 |
9 | 0:0 | 0:1 | 0:1 | 0:1 | 0:1 | 0:1 | 0:1 | 0:1 | 0:1 |
The dataset is too sparse and small, partly due to error of unknown origin outside the control of the student.
In order to know at what ages which age differences are problematic
Thank you for listening