{"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/f-vaegan-d2-a-feature-generating-framework","title":"f-VAEGAN-D2: A Feature Generating Framework for Any-Shot Learning","arxiv_id":"1903.10132","date":"2019-03-25","proceeding":"CVPR 2019 6","authors":["Yongqin Xian","Saurabh Sharma","Bernt Schiele","Zeynep Akata"],"abstract":"When labeled training data is scarce, a promising data augmentation approach\nis to generate visual features of unknown classes using their attributes. To\nlearn the class conditional distribution of CNN features, these models rely on\npairs of image features and class attributes. Hence, they can not make use of\nthe abundance of unlabeled data samples. In this paper, we tackle any-shot\nlearning problems i.e. zero-shot and few-shot, in a unified feature generating\nframework that operates in both inductive and transductive learning settings.\nWe develop a conditional generative model that combines the strength of VAE and\nGANs and in addition, via an unconditional discriminator, learns the marginal\nfeature distribution of unlabeled images. We empirically show that our model\nlearns highly discriminative CNN features for five datasets, i.e. CUB, SUN, AWA\nand ImageNet, and establish a new state-of-the-art in any-shot learning, i.e.\ninductive and transductive (generalized) zero- and few-shot learning settings.\nWe also demonstrate that our learned features are interpretable: we visualize\nthem by inverting them back to the pixel space and we explain them by\ngenerating textual arguments of why they are associated with a certain label.","url_abs":"http://arxiv.org/abs/1903.10132v1","url_pdf":"http://arxiv.org/pdf/1903.10132v1.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":[],"tasks":[{"task_slug":"data-augmentation","task_name":"Data Augmentation"},{"task_slug":"few-shot-learning","task_name":"Few-Shot Learning"},{"task_slug":"generalized-zero-shot-learning","task_name":"Generalized Zero-Shot Learning"},{"task_slug":"transductive-learning","task_name":"Transductive Learning"},{"task_slug":"zero-shot-learning","task_name":"Zero-Shot Learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/generalized-zero-shot-learning-on-sun","task":"Generalized Zero-Shot Learning","dataset":"SUN Attribute","model":"f-VAEGAN","rank_in_archive_order":4,"of":9,"metrics":{"Harmonic mean":"41.3"},"uses_additional_data":false},{"leaderboard":"/sota/zero-shot-learning-on-cub-200-2011","task":"Zero-Shot Learning","dataset":"CUB-200-2011","model":"f-VAEGAN-D2","rank_in_archive_order":8,"of":14,"metrics":{"average top-1 classification accuracy":"61.0"},"uses_additional_data":false},{"leaderboard":"/sota/zero-shot-learning-on-sun-attribute","task":"Zero-Shot Learning","dataset":"SUN Attribute","model":"f-VAEGAN","rank_in_archive_order":4,"of":9,"metrics":{"average top-1 classification accuracy":"64.7"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1903.10132","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}