{"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/unsupervised-embedding-learning-via-invariant","title":"Unsupervised Embedding Learning via Invariant and Spreading Instance Feature","arxiv_id":"1904.03436","date":"2019-04-06","proceeding":"CVPR 2019 6","authors":["Mang Ye","Xu Zhang","Pong C. Yuen","Shih-Fu Chang"],"abstract":"This paper studies the unsupervised embedding learning problem, which\nrequires an effective similarity measurement between samples in low-dimensional\nembedding space. Motivated by the positive concentrated and negative separated\nproperties observed from category-wise supervised learning, we propose to\nutilize the instance-wise supervision to approximate these properties, which\naims at learning data augmentation invariant and instance spread-out features.\nTo achieve this goal, we propose a novel instance based softmax embedding\nmethod, which directly optimizes the `real' instance features on top of the\nsoftmax function. It achieves significantly faster learning speed and higher\naccuracy than all existing methods. The proposed method performs well for both\nseen and unseen testing categories with cosine similarity. It also achieves\ncompetitive performance even without pre-trained network over samples from\nfine-grained categories.","url_abs":"http://arxiv.org/abs/1904.03436v1","url_pdf":"http://arxiv.org/pdf/1904.03436v1.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":"unsupervised-embedding-learning-via-invariant","repo_url":"https://github.com/mangye16/Unsupervised_Embedding_Learning","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"data-augmentation","task_name":"Data Augmentation"}],"methods":[{"method_slug":"speed","method_name":"SPEED"},{"method_slug":"softmax","method_name":"Softmax"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1904.03436","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}