{"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/class2str-end-to-end-latent-hierarchy","title":"Class2Str: End to End Latent Hierarchy Learning","arxiv_id":"1808.06675","date":"2018-08-20","proceeding":null,"authors":["Soham Saha","Girish Varma","C. V. Jawahar"],"abstract":"Deep neural networks for image classification typically consists of a\nconvolutional feature extractor followed by a fully connected classifier\nnetwork. The predicted and the ground truth labels are represented as one hot\nvectors. Such a representation assumes that all classes are equally dissimilar.\nHowever, classes have visual similarities and often form a hierarchy. Learning\nthis latent hierarchy explicitly in the architecture could provide invaluable\ninsights. We propose an alternate architecture to the classifier network called\nthe Latent Hierarchy (LH) Classifier and an end to end learned Class2Str\nmapping which discovers a latent hierarchy of the classes. We show that for\nsome of the best performing architectures on CIFAR and Imagenet datasets, the\nproposed replacement and training by LH classifier recovers the accuracy, with\na fraction of the number of parameters in the classifier part. Compared to the\nprevious work of HDCNN, which also learns a 2 level hierarchy, we are able to\nlearn a hierarchy at an arbitrary number of levels as well as obtain an\naccuracy improvement on the Imagenet classification task over them. We also\nverify that many visually similar classes are grouped together, under the\nlearnt hierarchy.","url_abs":"http://arxiv.org/abs/1808.06675v1","url_pdf":"http://arxiv.org/pdf/1808.06675v1.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":"class2str-end-to-end-latent-hierarchy","repo_url":"https://github.com/Soham0/Class2Str","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"image-classification","task_name":"image-classification"}],"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}