{"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/weakly-supervised-soft-detection-based","title":"Adversarial Soft-detection-based Aggregation Network for Image Retrieval","arxiv_id":"1811.07619","date":"2018-11-19","proceeding":null,"authors":["Jian Xu","Chunheng Wang","Cunzhao Shi","Baihua Xiao"],"abstract":"In recent year, the compact representations based on activations of\nConvolutional Neural Network (CNN) achieve remarkable performance in image\nretrieval. However, retrieval of some interested object that only takes up a\nsmall part of the whole image is still a challenging problem. Therefore, it is\nsignificant to extract the discriminative representations that contain regional\ninformation of the pivotal small object. In this paper, we propose a novel\nadversarial soft-detection-based aggregation (ASDA) method free from bounding\nbox annotations for image retrieval, based on adversarial detector and soft\nregion proposal layer. Our trainable adversarial detector generates semantic\nmaps based on adversarial erasing strategy to preserve more discriminative and\ndetailed information. Computed based on semantic maps corresponding to various\ndiscriminative patterns and semantic contents, our soft region proposal is\narbitrary shape rather than only rectangle and it reflects the significance of\nobjects. The aggregation based on trainable soft region proposal highlights\ndiscriminative semantic contents and suppresses the noise of background.\n  We conduct comprehensive experiments on standard image retrieval datasets.\nOur weakly supervised ASDA method achieves state-of-the-art performance on most\ndatasets. The results demonstrate that the proposed ASDA method is effective\nfor image retrieval.","url_abs":"http://arxiv.org/abs/1811.07619v3","url_pdf":"http://arxiv.org/pdf/1811.07619v3.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":"weakly-supervised-soft-detection-based","repo_url":"https://github.com/hbwang1427/image_retrieval","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}}],"tasks":[{"task_slug":"image-retrieval","task_name":"Image Retrieval"},{"task_slug":"region-proposal","task_name":"Region Proposal"},{"task_slug":"retrieval","task_name":"Retrieval"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}