Doucene, Salah EddineZehani, Mohamed HachedGhemougui, Abdessattar: Reporter2024-06-302024-06-302024-06-12https://repository.univ-msila.dz/handle/123456789/42917Age, gender, and ethnicity recognition technology analyzes faces using computer vision, but accuracy can be impacted by lighting, pose, and image quality. This report investigates existing methods, trains deep models similar to VGG/ResNet architectures and a pre-trained model from SkillCate, then evaluates and discusses the results. It aims to improve recognition accuracy despite these challenges.enHUMANITIES and RELIGION::Languages and linguistics::Scandinavian languages::Norwegian languageGender and Ethnicity RecognitionComputer VisionThesis