Layer-by-Layer Knowledge Distillation for Training Simplified Bipolar Morphological Neural Networks

Various neuron approximations can be used to reduce the computational complexity of neural networks. One such approximation based on summation and maximum operations is a bipolar morphological neuron. This paper presents an improved structure of the bipolar morphological neuron that enhances its computational efficiency and a new approach to training based on continuous approximations of the maximum and knowledge distillation. Experiments were carried out on the MNIST dataset using a LeNet-like neural network architecture and on the CIFAR10 dataset using a ResNet-22 model architecture. The proposed training method achieves 99.45% classification accuracy on the LeNet-like model (the same accuracy as that provided by the classical network) and 86.69% accuracy on the ResNet-22 model compared with 86.43% accuracy of the classical model. The results show that the proposed method with log-sum-exp (LSE) approximation of the maximum and layer-by-layer knowledge distillation makes it possible to obtain a simplified bipolar morphological network that is not inferior to the classical networks.