The Multi-Lane Capsule Network (MLCN)
This work addresses efficiency issues in Capsule Networks for researchers and practitioners, though it appears incremental as it builds on existing CapsNet methods.
The paper tackles the computational cost and efficiency of Capsule Networks by introducing Multi-Lane Capsule Networks (MLCN), which achieve similar accuracy with a much reduced number of parameters and more than two-fold faster training and inference times on datasets like Fashion-MNIST and Cifar10.
We introduce Multi-Lane Capsule Networks (MLCN), which are a separable and resource efficient organization of Capsule Networks (CapsNet) that allows parallel processing, while achieving high accuracy at reduced cost. A MLCN is composed of a number of (distinct) parallel lanes, each contributing to a dimension of the result, trained using the routing-by-agreement organization of CapsNet. Our results indicate similar accuracy with a much reduced cost in number of parameters for the Fashion-MNIST and Cifar10 datsets. They also indicate that the MLCN outperforms the original CapsNet when using a proposed novel configuration for the lanes. MLCN also has faster training and inference times, being more than two-fold faster than the original CapsNet in the same accelerator.