{"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-convolutional-neural-networks","title":"Quaternion Convolutional Neural Networks","arxiv_id":"1903.00658","date":"2019-03-02","proceeding":"ECCV 2018 9","authors":["Xuanyu Zhu","Yi Xu","Hongteng Xu","Changjian Chen"],"abstract":"Neural networks in the real domain have been studied for a long time and\nachieved promising results in many vision tasks for recent years. However, the\nextensions of the neural network models in other number fields and their\npotential applications are not fully-investigated yet. Focusing on color\nimages, which can be naturally represented as quaternion matrices, we propose a\nquaternion convolutional neural network (QCNN) model to obtain more\nrepresentative features. In particular, we redesign the basic modules like\nconvolution layer and fully-connected layer in the quaternion domain, which can\nbe used to establish fully-quaternion convolutional neural networks. Moreover,\nthese modules are compatible with almost all deep learning techniques and can\nbe plugged into traditional CNNs easily. We test our QCNN models in both color\nimage classification and denoising tasks. Experimental results show that they\noutperform the real-valued CNNs with same structures.","url_abs":"http://arxiv.org/abs/1903.00658v1","url_pdf":"http://arxiv.org/pdf/1903.00658v1.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-convolutional-neural-networks","repo_url":"https://github.com/jorisweeda/quaternion-convolutional-neural-networks","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"quaternion-convolutional-neural-networks","repo_url":"https://github.com/XYZ387/QuaternionCNN_Keras","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"tf","reach":{"status":"ok"}}],"tasks":[{"task_slug":"denoising","task_name":"Denoising"},{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"image-classification","task_name":"image-classification"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1903.00658","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}