Convolutional Neural Network Compression through Generalized Kronecker Product Decomposition

Modern Convolutional Neural Network (CNN) architectures, despite their superiority in solving various problems, are generally too large to be deployed on resource constrained edge devices. In this paper, we reduce memory usage and floating-point operations required by convolutional layers CNNs. We compress these generalizing the Kronecker Product Decomposition apply multidimensional tensors, leading Generalized (GKPD). Our approach yields a plug-and-play module that can used as drop-in replacement for any layer. Experimental results image classification CIFAR-10 ImageNet datasets using ResNet, MobileNetv2 SeNet architectures substantiate effectiveness of our proposed approach. find GKPD outperforms state-of-the-art decomposition methods including Tensor-Train Tensor-Ring well other relevant compression such pruning knowledge distillation.