Towards Efficient Capsule Networks
This work addresses the problem of deploying Capsule Networks on resource-constrained devices for more complex tasks, though it is incremental as it applies existing pruning strategies to a specific network type.
The paper tackles the computational inefficiency and low scalability of Capsule Networks by introducing sparsity through pruning to reduce the number of capsules, resulting in improved generalization with reduced memory, computational effort, and inference and training time.
From the moment Neural Networks dominated the scene for image processing, the computational complexity needed to solve the targeted tasks skyrocketed: against such an unsustainable trend, many strategies have been developed, ambitiously targeting performance's preservation. Promoting sparse topologies, for example, allows the deployment of deep neural networks models on embedded, resource-constrained devices. Recently, Capsule Networks were introduced to enhance explainability of a model, where each capsule is an explicit representation of an object or its parts. These models show promising results on toy datasets, but their low scalability prevents deployment on more complex tasks. In this work, we explore sparsity besides capsule representations to improve their computational efficiency by reducing the number of capsules. We show how pruning with Capsule Network achieves high generalization with less memory requirements, computational effort, and inference and training time.