“…Biologically-inspired techniques have also been adopted by using a feed-forward neural network (NN) [27] to achieve background subtraction. A recent effort based on the NN [28] extends [29] the idea of self-organisation to automatically compensate for the egomotion of a device. Convolutional NNs are used recently for background subtraction where the background modelling is learned using the hand segmented foreground objects in the training images [30].…”