Fig. 4From: Viewpoint robust knowledge distillation for accelerating vehicle re-identificationVisualizations of features resulted from SDC-CNN-based student networks using a VRKD and not b using VRKD, respectivelyBack to article page