Quaternion Generative Adversarial Networks

04/19/2021
by   Eleonora Grassucci, et al.
0

Latest Generative Adversarial Networks (GANs) are gathering outstanding results through a large-scale training, thus employing models composed of millions of parameters requiring extensive computational capabilities. Building such huge models undermines their replicability and increases the training instability. Moreover, multi-channel data, such as images or audio, are usually processed by real-valued convolutional networks that flatten and concatenate the input, losing any intra-channel spatial relation. To address these issues, here we propose a family of quaternion-valued generative adversarial networks (QGANs). QGANs exploit the properties of quaternion algebra, e.g., the Hamilton product for convolutions. This allows to process channels as a single entity and capture internal latent relations, while reducing by a factor of 4 the overall number of parameters. We show how to design QGANs and to extend the proposed approach even to advanced models. We compare the proposed QGANs with real-valued counterparts on multiple image generation benchmarks. Results show that QGANs are able to generate visually pleasing images and to obtain better FID scores with respect to their real-valued GANs. Furthermore, QGANs save up to 75 novel, more accessible, GANs capable of improving performance and saving computational resources.

READ FULL TEXT

page 22

page 23

research
12/05/2017

Manifold-valued Image Generation with Wasserstein Adversarial Networks

Unsupervised image generation has recently received an increasing amount...
research
04/25/2018

Convolutional Generative Adversarial Networks with Binary Neurons for Polyphonic Music Generation

It has been shown recently that convolutional generative adversarial net...
research
09/09/2020

Generalizing Complex/Hyper-complex Convolutions to Vector Map Convolutions

We show that the core reasons that complex and hypercomplex valued neura...
research
09/20/2022

Review of data types and model dimensionality for cardiac DTI SMS-related artefact removal

As diffusion tensor imaging (DTI) gains popularity in cardiac imaging du...
research
08/03/2020

Generative Adversarial Networks for Synthesizing InSAR Patches

Generative Adversarial Networks (GANs) have been employed with certain s...
research
10/26/2022

Anisotropic multiresolution analyses for deep fake detection

Generative Adversarial Networks (GANs) have paved the path towards entir...
research
10/04/2021

GenCo: Generative Co-training on Data-Limited Image Generation

Training effective Generative Adversarial Networks (GANs) requires large...

Please sign up or login with your details

Forgot password? Click here to reset