Person re-identification re-ID (re-ID) aim is to match the same person across disjoint camera views. Given a query image, the goal of re-ID is retrieving images in the gallery images. In the person re-ID, it is important to be aware of subtle ID features such as body shape, hairstyle, or face appearance. However, because of different viewpoints, occlusion, or illumination changes, person re-ID remains a challenging task.
Knowledge distillation refers to the method that helps the training process of a smaller network (student) under the supervision of a larger network (teacher). Unlike other compression methods, it can downsize a network regardless of the structural difference between the teacher and the student network