{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/quaternion-generative-adversarial-networks","title":"Quaternion Generative Adversarial Networks","arxiv_id":"2104.09630","date":"2021-04-19","proceeding":null,"authors":["Eleonora Grassucci","Edoardo Cicero","Danilo Comminiello"],"abstract":"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 realvalued convolutional networks that flatten and concatenate the input, often losing intra-channel spatial relations. To address these issues related to complexity and information loss, we propose a family of quaternion-valued generative adversarial networks (QGANs). QGANs exploit the properties of quaternion algebra, e.g., the Hamilton product, that 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 several image generation benchmarks. Results show that QGANs are able to obtain better FID scores than real-valued GANs and to generate visually pleasing images. Furthermore, QGANs save up to 75% of the training parameters. We believe these results may pave the way to novel, more accessible, GANs capable of improving performance and saving computational resources.","url_abs":"https://arxiv.org/abs/2104.09630v2","url_pdf":"https://arxiv.org/pdf/2104.09630v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"quaternion-generative-adversarial-networks","repo_url":"https://github.com/eleGAN23/QGAN","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"quaternion-generative-adversarial-networks","repo_url":"https://github.com/eleGAN23/QVAE","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"quaternion-generative-adversarial-networks","repo_url":"https://github.com/ispamm/hi2i","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"image-generation","task_name":"Image Generation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/image-generation-on-cifar-10","task":"Image Generation","dataset":"CIFAR-10","model":"QSNGAN","rank_in_archive_order":69,"of":78,"metrics":{"FID":"31.966"},"uses_additional_data":false},{"leaderboard":"/sota/image-generation-on-celeba-hq-128x128","task":"Image Generation","dataset":"CelebA-HQ 128x128","model":"QSNGAN","rank_in_archive_order":7,"of":7,"metrics":{"FID":"29.417","IS":"2.249"},"uses_additional_data":false},{"leaderboard":"/sota/image-generation-on-oxford-102-flowers-1","task":"Image Generation","dataset":"Oxford 102 Flowers 128x128","model":"QSNGAN","rank_in_archive_order":1,"of":1,"metrics":{"FID":"115.838","IS":"3"},"uses_additional_data":false},{"leaderboard":"/sota/image-generation-on-stl-10","task":"Image Generation","dataset":"STL-10","model":"QSNGAN","rank_in_archive_order":30,"of":31,"metrics":{"FID":"59.611","Inception score":"4.987"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/2104.09630","atlas_url":"https://app.syntology.ai/?focus=2104.09630","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}