{"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/low-shot-visual-recognition-by-shrinking-and","title":"Low-shot Visual Recognition by Shrinking and Hallucinating Features","arxiv_id":"1606.02819","date":"2016-06-09","proceeding":"ICCV 2017 10","authors":["Bharath Hariharan","Ross Girshick"],"abstract":"Low-shot visual learning---the ability to recognize novel object categories\nfrom very few examples---is a hallmark of human visual intelligence. Existing\nmachine learning approaches fail to generalize in the same way. To make\nprogress on this foundational problem, we present a low-shot learning benchmark\non complex images that mimics challenges faced by recognition systems in the\nwild. We then propose a) representation regularization techniques, and b)\ntechniques to hallucinate additional training examples for data-starved\nclasses. Together, our methods improve the effectiveness of convolutional\nnetworks in low-shot learning, improving the one-shot accuracy on novel classes\nby 2.3x on the challenging ImageNet dataset.","url_abs":"http://arxiv.org/abs/1606.02819v4","url_pdf":"http://arxiv.org/pdf/1606.02819v4.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":"low-shot-visual-recognition-by-shrinking-and","repo_url":"https://github.com/facebookresearch/low-shot-shrink-hallucinate","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"low-shot-visual-recognition-by-shrinking-and","repo_url":"https://github.com/MyChocer/KGTN","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"low-shot-visual-recognition-by-shrinking-and","repo_url":"https://github.com/RajeevReddyIlavala/Low-Shot-Learning-Learning","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"NOASSERTION"}},{"paper_slug":"low-shot-visual-recognition-by-shrinking-and","repo_url":"https://github.com/sambi97/Low-shot-learning_-Deep-learning-project","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"NOASSERTION"}}],"tasks":[{"task_slug":"machine-learning","task_name":"BIG-bench Machine Learning"},{"task_slug":"few-shot-image-classification","task_name":"Few-Shot Image Classification"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/few-shot-image-classification-on-imagenet-fs","task":"Few-Shot Image Classification","dataset":"ImageNet-FS (1-shot, novel)","model":"SGM (ResNet-50)","rank_in_archive_order":7,"of":7,"metrics":{"Top-5 Accuracy (%)":"52.9"},"uses_additional_data":false},{"leaderboard":"/sota/few-shot-image-classification-on-imagenet-fs-1","task":"Few-Shot Image Classification","dataset":"ImageNet-FS (2-shot, novel)","model":"SGM (ResNet-50)","rank_in_archive_order":6,"of":8,"metrics":{"Top-5 Accuracy (%)":"67.0"},"uses_additional_data":false},{"leaderboard":"/sota/few-shot-image-classification-on-imagenet-fs-1","task":"Few-Shot Image Classification","dataset":"ImageNet-FS (2-shot, novel)","model":"SGM w/G (ResNet-50)","rank_in_archive_order":8,"of":8,"metrics":{"Top-5 Accuracy (%)":"64.9"},"uses_additional_data":false},{"leaderboard":"/sota/few-shot-image-classification-on-imagenet-fs-6","task":"Few-Shot Image Classification","dataset":"ImageNet-FS (5-shot, all)","model":"SGM (ResNet-50)","rank_in_archive_order":5,"of":8,"metrics":{"Top-5 Accuracy (%)":"77.4"},"uses_additional_data":false},{"leaderboard":"/sota/few-shot-image-classification-on-imagenet-fs-6","task":"Few-Shot Image Classification","dataset":"ImageNet-FS (5-shot, all)","model":"SGM w/ G  (ResNet-50)","rank_in_archive_order":7,"of":8,"metrics":{"Top-5 Accuracy (%)":"77.3"},"uses_additional_data":false}],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}