Computational Intelligence and Neuroscience / 2018 / Article / Tab 1

Research Article

A Community Detection Approach to Cleaning Extremely Large Face Database

Table 1

Comparison with other methods in terms of cleaning MS-Celeb-1M. Ours is able to achieve high cleanness, rich data diversity, and large data scale at the same time.

MethodkρζCleannessDiversity#Subject#Image

Ours-5-93.6%0.581197,6467,244,505
-10-97.2%0.551391,1806,024,931
-15-98.3%0.544984,2975,738,961
-20-98.8%0.529378,3995,082,658

MSM---98.9%0.484399,8922,207,688

KCM55-68.2%0.723199,5447,644,424
10-70.1%0.704799,4797,177,587
15-73.2%0.676399,4266,291,849
20-74.0%0.638699,3845,119,874
75-68.0%0.716999,3047,500,139
10-72.0%0.681699,2276,428,188
15-77.5%0.633499,1454,892,615
20-82.9%0.567494,8123,359,597

FPR--0.267.7%0.660399,8926,225,585
--0.369.0%0.631999,8925,458,018
--0.475.1%0.603599,8924,679,810
--0.580.8%0.574699,8923,890,961
--0.5579.5%0.561099,8923,526,922
--0.680.2%0.545399,8923,133,376

LCNN---98.8%0.506279,0775,049,824

None---61.1%0.727799,8928,456,240

We are committed to sharing findings related to COVID-19 as quickly as possible. We will be providing unlimited waivers of publication charges for accepted research articles as well as case reports and case series related to COVID-19. Review articles are excluded from this waiver policy. Sign up here as a reviewer to help fast-track new submissions.