{"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/object-part-attention-model-for-fine-grained","title":"Object-Part Attention Model for Fine-grained Image Classification","arxiv_id":"1704.01740","date":"2017-04-06","proceeding":null,"authors":["Yuxin Peng","Xiangteng He","Junjie Zhao"],"abstract":"Fine-grained image classification is to recognize hundreds of subcategories\nbelonging to the same basic-level category, such as 200 subcategories belonging\nto the bird, which is highly challenging due to large variance in the same\nsubcategory and small variance among different subcategories. Existing methods\ngenerally first locate the objects or parts and then discriminate which\nsubcategory the image belongs to. However, they mainly have two limitations:\n(1) Relying on object or part annotations which are heavily labor consuming.\n(2) Ignoring the spatial relationships between the object and its parts as well\nas among these parts, both of which are significantly helpful for finding\ndiscriminative parts. Therefore, this paper proposes the object-part attention\nmodel (OPAM) for weakly supervised fine-grained image classification, and the\nmain novelties are: (1) Object-part attention model integrates two level\nattentions: object-level attention localizes objects of images, and part-level\nattention selects discriminative parts of object. Both are jointly employed to\nlearn multi-view and multi-scale features to enhance their mutual promotions.\n(2) Object-part spatial constraint model combines two spatial constraints:\nobject spatial constraint ensures selected parts highly representative, and\npart spatial constraint eliminates redundancy and enhances discrimination of\nselected parts. Both are jointly employed to exploit the subtle and local\ndifferences for distinguishing the subcategories. Importantly, neither object\nnor part annotations are used in our proposed approach, which avoids the heavy\nlabor consumption of labeling. Comparing with more than 10 state-of-the-art\nmethods on 4 widely-used datasets, our OPAM approach achieves the best\nperformance.","url_abs":"http://arxiv.org/abs/1704.01740v2","url_pdf":"http://arxiv.org/pdf/1704.01740v2.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":"object-part-attention-model-for-fine-grained","repo_url":"https://github.com/PKU-ICST-MIPL/OPAM_TIP2018","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"none","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"classification-1","task_name":"Classification"},{"task_slug":"fine-grained-image-classification","task_name":"Fine-Grained Image Classification"},{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"object","task_name":"Object"},{"task_slug":"image-classification","task_name":"image-classification"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1704.01740","atlas_url":"https://app.syntology.ai/?focus=1704.01740","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}