{"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/few-shot-learning-with-part-discovery-and","title":"Few-Shot Learning with Part Discovery and Augmentation from Unlabeled Images","arxiv_id":"2105.11874","date":"2021-05-25","proceeding":null,"authors":["Wentao Chen","Chenyang Si","Wei Wang","Liang Wang","Zilei Wang","Tieniu Tan"],"abstract":"Few-shot learning is a challenging task since only few instances are given for recognizing an unseen class. One way to alleviate this problem is to acquire a strong inductive bias via meta-learning on similar tasks. In this paper, we show that such inductive bias can be learned from a flat collection of unlabeled images, and instantiated as transferable representations among seen and unseen classes. Specifically, we propose a novel part-based self-supervised representation learning scheme to learn transferable representations by maximizing the similarity of an image to its discriminative part. To mitigate the overfitting in few-shot classification caused by data scarcity, we further propose a part augmentation strategy by retrieving extra images from a base dataset. We conduct systematic studies on miniImageNet and tieredImageNet benchmarks. Remarkably, our method yields impressive results, outperforming the previous best unsupervised methods by 7.74% and 9.24% under 5-way 1-shot and 5-way 5-shot settings, which are comparable with state-of-the-art supervised methods.","url_abs":"https://arxiv.org/abs/2105.11874v1","url_pdf":"https://arxiv.org/pdf/2105.11874v1.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":"few-shot-learning","task_name":"Few-Shot Learning"},{"task_slug":"inductive-bias","task_name":"Inductive Bias"},{"task_slug":"meta-learning","task_name":"Meta-Learning"},{"task_slug":"representation-learning","task_name":"Representation Learning"},{"task_slug":"unsupervised-few-shot-image-classification","task_name":"Unsupervised Few-Shot Image Classification"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/unsupervised-few-shot-image-classification-on","task":"Unsupervised Few-Shot Image Classification","dataset":"Mini-Imagenet 5-way (1-shot)","model":"PDA-Net","rank_in_archive_order":4,"of":28,"metrics":{"Accuracy":"63.84"},"uses_additional_data":false},{"leaderboard":"/sota/unsupervised-few-shot-image-classification-on-1","task":"Unsupervised Few-Shot Image Classification","dataset":"Mini-Imagenet 5-way (5-shot)","model":"PDA-Net","rank_in_archive_order":4,"of":28,"metrics":{"Accuracy":"83.11"},"uses_additional_data":false},{"leaderboard":"/sota/unsupervised-few-shot-image-classification-on-2","task":"Unsupervised Few-Shot Image Classification","dataset":"Tiered ImageNet 5-way (1-shot)","model":"PDA-Net","rank_in_archive_order":3,"of":12,"metrics":{"Accuracy":"69.01"},"uses_additional_data":false},{"leaderboard":"/sota/unsupervised-few-shot-image-classification-on-3","task":"Unsupervised Few-Shot Image Classification","dataset":"Tiered ImageNet 5-way (5-shot)","model":"PDA-Net","rank_in_archive_order":4,"of":12,"metrics":{"Accuracy":"84.20"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/2105.11874","atlas_url":"https://app.syntology.ai/?focus=2105.11874","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}