A Continuous Convolutional Trainable Filter for Modelling Unstructured Data

TL;DR

This paper introduces a continuous convolutional filter that extends CNN capabilities to unstructured data, enabling broader applications beyond traditional discrete domains while maintaining competitive accuracy.

Contribution

The paper presents a novel continuous trainable convolutional filter framework that works with unstructured data, expanding CNN applicability to more complex, non-grid data domains.

Findings

Achieves accuracy comparable to state-of-the-art discrete filters.

Enables CNNs to process unstructured data effectively.

Integrates seamlessly into existing deep learning architectures.

Abstract

Convolutional Neural Network (CNN) is one of the most important architectures in deep learning. The fundamental building block of a CNN is a trainable filter, represented as a discrete grid, used to perform convolution on discrete input data. In this work, we propose a continuous version of a trainable convolutional filter able to work also with unstructured data. This new framework allows exploring CNNs beyond discrete domains, enlarging the usage of this important learning technique for many more complex problems. Our experiments show that the continuous filter can achieve a level of accuracy comparable to the state-of-the-art discrete filter, and that it can be used in current deep learning architectures as a building block to solve problems with unstructured domains as well.