2 Timing Sample Dataset

A fixed sample of images randomly selected from the PFT III datasets are used to assess whether or not an implementation adheres to the minimum timing requirements set forth in the PFT III test plan. This sample is also used to provide an estimate on template size. The images and comparisons are identical to the “1K Sample Evaluation” from NIST’s PFT II evaluation, with the exception of the “IARPA N2N” dataset, which is new in PFT III. Table 2.1 shows information about the maximum dimensions and resolutions of the images in each of the timing sample datasets.

Table 2.1: Maximum dimensions in pixels and capture resolution in pixels per inch (PPI) for the images in each of the subsets comprising the timing sample dataset.
AZ LA County DHS 2 POE+BVA IARPA N2N
Max Dimensions (pixels) 800 x 800 412 x 1 000 368 x 368 500 x 500 1 600 x 1 500
Resolution (PPI) 500 500 500 500 1 000

2.1 Template Size

Figure 2.1 and Table 2.2 show the distribution of file sizes of templates. Failures of any kind reported during template generation result in NIST code writing 0 byte files. These files are excluded from the template size analysis in this section.

Box plots of template sizes in bytes of templates created from a fixed sample of data from the PFT III evaluation. An overall plot is shown, as well as individual plots per data origin. Tabular versions of this data are shown in Table 2.2.

Figure 2.1: Box plots of template sizes in bytes of templates created from a fixed sample of data from the PFT III evaluation. An overall plot is shown, as well as individual plots per data origin. Tabular versions of this data are shown in Table 2.2.

Table 2.2: Sizes in bytes of templates created from a fixed sample of data from the PFT III evaluation. The bottom row, Failures, shows the number of failures to create a template, which are not included in these statistics. Box plots of this data are shown in Figure 2.1.
Overall AZ LA County DHS 2 POE+BVA IARPA N2N
Min 2 624 12 200 2 624 2 890 2 624 10 870
25 % 15 924 28 692 25 500 13 796 13 530 28 692
Median 23 904 30 288 28 692 15 924 15 658 28 692
Mean 23 380 32 220 27 269 15 954 15 790 30 238
75 % 28 692 34 810 28 692 18 318 17 786 30 820
Max 69 922 55 558 46 780 28 426 28 692 69 922
Failures 2 0 1 1 0 0

2.2 Template Creation Time

Figure 2.2 and Table 2.3 show the distribution of durations of time consumed when creating templates. Failures of all kinds are incorporated into these statistics, since this time would be observed by the end user of a template creation algorithm.

Times are measured by running a single process on an isolated compute node equipped with an Intel Xeon Gold 6254 CPU.

Box plots of elapsed milliseconds when creating templates from a fixed sample of data from the PFT III evaluation. All times are used, even if a failure occurred. Tabular versions of this data are shown in Table 2.3.

Figure 2.2: Box plots of elapsed milliseconds when creating templates from a fixed sample of data from the PFT III evaluation. All times are used, even if a failure occurred. Tabular versions of this data are shown in Table 2.3.

Table 2.3: Elapsed milliseconds when creating templates from a fixed sample of data from the PFT III evaluation. All times are used, even if a failure occurred. The bottom row, Failures, shows the number of failures to generate a template. Failures are included in these statistics. Box plots of this data are shown in Figure 2.2.
Overall AZ LA County DHS 2 POE+BVA IARPA N2N
Min 101.0 291.7 137.1 101.0 110.1 266.6
25 % 210.6 427.3 331.5 188.9 188.6 410.8
Median 317.6 440.5 348.5 206.3 209.7 412.2
Mean 314.1 455.8 344.0 206.8 213.1 424.7
75 % 418.3 476.1 362.9 225.5 234.5 430.2
Max 747.4 643.1 494.2 306.6 333.5 747.4
Failures 2 0 1 1 0 0

2.3 Template Comparison Time

Figure 2.3 and Table 2.4 show the distribution of durations of time consumed when comparing templates. Failures of any kind are incorporated into these statistics, since this time would be observed by the end user of a template comparison algorithm.

Times are measured by running a single process on an isolated compute node equipped with an Intel Xeon Gold 6254 CPU.

Box plots of elapsed microseconds when comparing two templates from a fixed sample of data from the PFT III evaluation. All times are used, even if a failure occurred. Tabular versions of this data are shown in Table 2.4.

Figure 2.3: Box plots of elapsed microseconds when comparing two templates from a fixed sample of data from the PFT III evaluation. All times are used, even if a failure occurred. Tabular versions of this data are shown in Table 2.4.

Table 2.4: Elapsed microseconds when comparing two templates from a fixed sample of data from the PFT III evaluation. The bottom row, Failures, shows the number of failures to compare. Failures are included in these statistics. Box plots of this data are shown in Figure 2.4.
Overall AZ LA County DHS 2 POE+BVA IARPA N2N
Min 0.5 612.7 0.5 0.6 67.8 397.0
25 % 838.8 1 731.2 1 295.7 500.0 502.5 1 360.4
Median 1 391.3 2 231.4 1 705.3 677.2 799.0 1 533.8
Mean 1 650.6 2 751.1 2 013.7 756.2 972.9 1 761.0
75 % 2 044.4 3 283.9 2 391.1 935.4 1 198.8 1 969.8
Max 27 907.7 27 907.7 7 028.7 4 076.2 4 578.1 11 266.5
Failures 0 0 0 0 0 0