One of the best promising solutions to mitigate the negative effects of the mechanical wet separation and dry separation techniques that are used in Coal Preparation Plant is computer vision (CV), which is a field of artificial intelligence (AI), and the use of digital cameras or infrared cameras, or both of them, for coal and gangue recognition; then, the gangue can be removed from the production line, as shown in Figure. Although this system appears to have a complicated operation, it has many benefits (e.g., environmentally friendly processes, water conservation, safety, high sustainability, and low maintenance costs due to decreasing the rock content, which helps to reduce the wearing of the crusher teeth). These aspects prompted me to focus in depth on designing a deep learning algorithm for binary classification so that the performance of automated dry separation systems based on infrared machine vision can be improved.