Abstract: Presentation of a study and implementation on task named camouflage clothes detection. The aim of the study is to identify clothes that "seamlessly" blend in with their surroundings. Usage of a dataset, called 8k_normal_vs_camouflage_clothes_images, which consists of about 8k images of camouflage clothes and 8k images of normal clothes. Residual Networks (ResNets) are used for this task. The high intrinsic similarities between background and object make camouflage detection a challenging task