{"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/selective-convolutional-descriptor","title":"Selective Convolutional Descriptor Aggregation for Fine-Grained Image Retrieval","arxiv_id":"1604.04994","date":"2016-04-18","proceeding":null,"authors":["Xiu-Shen Wei","Jian-Hao Luo","Jianxin Wu","Zhi-Hua Zhou"],"abstract":"Deep convolutional neural network models pre-trained for the ImageNet\nclassification task have been successfully adopted to tasks in other domains,\nsuch as texture description and object proposal generation, but these tasks\nrequire annotations for images in the new domain. In this paper, we focus on a\nnovel and challenging task in the pure unsupervised setting: fine-grained image\nretrieval. Even with image labels, fine-grained images are difficult to\nclassify, let alone the unsupervised retrieval task. We propose the Selective\nConvolutional Descriptor Aggregation (SCDA) method. SCDA firstly localizes the\nmain object in fine-grained images, a step that discards the noisy background\nand keeps useful deep descriptors. The selected descriptors are then aggregated\nand dimensionality reduced into a short feature vector using the best practices\nwe found. SCDA is unsupervised, using no image label or bounding box\nannotation. Experiments on six fine-grained datasets confirm the effectiveness\nof SCDA for fine-grained image retrieval. Besides, visualization of the SCDA\nfeatures shows that they correspond to visual attributes (even subtle ones),\nwhich might explain SCDA's high mean average precision in fine-grained\nretrieval. Moreover, on general image retrieval datasets, SCDA achieves\ncomparable retrieval results with state-of-the-art general image retrieval\napproaches.","url_abs":"http://arxiv.org/abs/1604.04994v2","url_pdf":"http://arxiv.org/pdf/1604.04994v2.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":"selective-convolutional-descriptor","repo_url":"https://github.com/luiscarlosgph/videosum","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"image-retrieval","task_name":"Image Retrieval"},{"task_slug":"object-proposal-generation","task_name":"Object Proposal Generation"},{"task_slug":"retrieval","task_name":"Retrieval"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1604.04994","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}