Quaternion Convolutional Neural Networks: Current Advances and Future Directions

Abstract

Since their first applications, Convolutional Neural Networks (CNNs) have solved problems that have advanced the state-of-the-art in several domains. CNNs represent information using real numbers. Despite encouraging results, theoretical analysis shows that representations such as hyper-complex numbers can achieve richer representational capacities than real numbers, and that Hamilton products can capture intrinsic interchannel relationships. Moreover, in the last few years, experimental research has shown that Quaternion-Valued CNNs (QCNNs) can achieve similar performance with fewer parameters than their real-valued counterparts. This paper condenses research in the development of QCNNs from its very beginnings. We propose a conceptual organization of current trends and analyze the main building blocks used in the design of QCNN models. Based on this conceptual organization, we propose future directions of research.

References in corpus (13)

• A Survey of the Recent Architectures of Deep Convolutional Neural Networks
• Sound Event Localization and Detection of Overlapping Sources Using Convolutional Recurrent Neural Networks
• Optimal whitening and decorrelation
• Quaternion Gradient and Hessian
• The Theory of Quaternion Matrix Derivatives
• Quaternion Derivatives: The GHR Calculus
• Quaternion Convolutional Neural Networks for Heterogeneous Image Processing
• Fast complexified quaternion Fourier transform
• Quaternion Convolutional Neural Networks for Detection and Localization of 3D Sound Events
• Quaternion polar representation with a complex modulus and complex argument inspired by the Cayley-Dickson form
• Acute Lymphoblastic Leukemia Detection Using Hypercomplex-Valued Convolutional Neural Networks
• Calculus with a Quaternionic Variable
• Hypercomplex Image-to-Image Translation