{"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/hitnet-a-neural-network-with-capsules","title":"HitNet: a neural network with capsules embedded in a Hit-or-Miss layer, extended with hybrid data augmentation and ghost capsules","arxiv_id":"1806.06519","date":"2018-06-18","proceeding":null,"authors":["Adrien Deliège","Anthony Cioppa","Marc Van Droogenbroeck"],"abstract":"Neural networks designed for the task of classification have become a\ncommodity in recent years. Many works target the development of better\nnetworks, which results in a complexification of their architectures with more\nlayers, multiple sub-networks, or even the combination of multiple classifiers.\nIn this paper, we show how to redesign a simple network to reach excellent\nperformances, which are better than the results reproduced with CapsNet on\nseveral datasets, by replacing a layer with a Hit-or-Miss layer. This layer\ncontains activated vectors, called capsules, that we train to hit or miss a\ncentral capsule by tailoring a specific centripetal loss function. We also show\nhow our network, named HitNet, is capable of synthesizing a representative\nsample of the images of a given class by including a reconstruction network.\nThis possibility allows to develop a data augmentation step combining\ninformation from the data space and the feature space, resulting in a hybrid\ndata augmentation process. In addition, we introduce the possibility for\nHitNet, to adopt an alternative to the true target when needed by using the new\nconcept of ghost capsules, which is used here to detect potentially mislabeled\nimages in the training data.","url_abs":"http://arxiv.org/abs/1806.06519v1","url_pdf":"http://arxiv.org/pdf/1806.06519v1.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":"hitnet-a-neural-network-with-capsules","repo_url":"https://github.com/bakirillov/capsules","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"data-augmentation","task_name":"Data Augmentation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}