A Comprehensive Survey on Vitiligo Detection Using Deep Learning

The survey discusses the challenges and advancements in using deep learning for vitiligo detection, highlighting the limitations of current datasets, particularly their size and lack of diversity in skin tones. It compares traditional machine learning methods with modern convolutional neural networks and segmentation architectures, noting the need for more comprehensive datasets and consistent image capture conditions. The paper emphasizes the importance of improving pre-processing and evaluation practices to develop more reliable and universally applicable vitiligo detection systems.