| Summary: | Guo <italic>et al.</italic> proposed a method for automated face-sketch recognition in their paper “Domain Alignment Embedding Network for Sketch Face Recognition”, where it is deemed to be superior to several other methods published in literature. However, the employed evaluation methodology contains several critical flaws, such that definitive conclusions with regards to performance comparisons between the methods considered cannot be made. Moreover, erroneous results are reported for the works published in literature given the employed evaluation framework, potentially compromising the accuracy of future works that use the results published by Guo <italic>et al.</italic> Discussions on these and other observations are given in this manuscript.
|
|---|