{"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/capsgan-using-dynamic-routing-for-generative","title":"CapsGAN: Using Dynamic Routing for Generative Adversarial Networks","arxiv_id":"1806.03968","date":"2018-06-07","proceeding":null,"authors":["Raeid Saqur","Sal Vivona"],"abstract":"In this paper, we propose a novel technique for generating images in the 3D\ndomain from images with high degree of geometrical transformations. By\ncoalescing two popular concurrent methods that have seen rapid ascension to the\nmachine learning zeitgeist in recent years: GANs (Goodfellow et. al.) and\nCapsule networks (Sabour, Hinton et. al.) - we present: \\textbf{CapsGAN}. We\nshow that CapsGAN performs better than or equal to traditional CNN based GANs\nin generating images with high geometric transformations using rotated MNIST.\nIn the process, we also show the efficacy of using capsules architecture in the\nGANs domain. Furthermore, we tackle the Gordian Knot in training GANs - the\nperformance control and training stability by experimenting with using\nWasserstein distance (gradient clipping, penalty) and Spectral Normalization.\nThe experimental findings of this paper should propel the application of\ncapsules and GANs in the still exciting and nascent domain of 3D image\ngeneration, and plausibly video (frame) generation.","url_abs":"http://arxiv.org/abs/1806.03968v1","url_pdf":"http://arxiv.org/pdf/1806.03968v1.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":"capsgan-using-dynamic-routing-for-generative","repo_url":"https://github.com/raeidsaqur/CapsGAN","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"image-generation","task_name":"Image Generation"},{"task_slug":"rotated-mnist","task_name":"Rotated MNIST"}],"methods":[{"method_slug":"spectral-normalization","method_name":"Spectral Normalization"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}