Comprehensive insights into evaluation and benchmarking of real-time skin detectors: Review, open issues & challenges, and recommended solutions

Evaluation and benchmarking of real-time skin detectors remain challenging because of multiple evaluation attributes that must be considered. Numerous evaluation and benchmarking techniques have been proposed, but they exhibit several limitations. Fixing multiple attributes based on benchmarking approaches by using other attributes limits reliable real-time skin detection. This paper presents comprehensive insights into the evaluation and benchmarking of real-time skin detectors on the basis of two critical directions. Current evaluation criteria highlight conflicting issues and benchmarking techniques to identify weak points, and possible solutions are discussed. The findings are as follows: (1) open issues and challenges to evaluation and benchmarking are emphasized; and (2) decision making using multiple criteria such as reliability, time complexity, and error rate within a dataset is used for evaluating and benchmarking real-time skin detectors to come up with solutions for future directions.