Comments on the “Kinship Face in the Wild” Data Sets

摘要: The Kinship Face in the Wild data sets, recently published TPAMI, are currently used as a benchmark for evaluation of kinship verification algorithms. We recommend that these sets no longer research unless there is compelling reason takes into account nature images. note most image pairs cropped from same photographs. Exploiting this cropping information, competitive but biased performance can be obtained using simple scoring approach, taking only rather than any features about kin information. To illustrate our motives, we provide classification results utilizing method based on similarity both images pair. Using simply distance chrominance averages Lab color space without training or specific features, achieve comparable to state-of-the-art methods. source code prove validity claims and ensure repeatability experiments.