{"corpus_id":4309276,"paper_sha":"18c125ce0f64e85577f7d30132cf0e92ec664bf4","doi":"10.1109/CVPR.2014.461","arxiv_id":"1311.3618","pmid":null,"pmcid":null,"mag_id":2047643928,"dblp_id":"conf/cvpr/CimpoiMKMV14","acl_id":null,"title":"Describing Textures in the Wild","year":2013,"publication_date":"2013-11-14","venue":"2014 IEEE Conference on Computer Vision and Pattern Recognition","journal":{"name":"2014 IEEE Conference on Computer Vision and Pattern Recognition","pages":"3606-3613","volume":null},"journal_issn":null,"journal_title":null,"publication_types":["JournalArticle","Conference"],"pubmed_pub_types":null,"s2_fields_of_study":["Computer Science","Environmental Science"],"reference_count":45,"citation_count":3364,"influential_citation_count":407,"is_open_access":true,"arxiv_categories":["cs.CV"],"arxiv_license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","arxiv_journal_ref":null,"mesh_headings":null,"chemicals":null,"comments_corrections":null,"source_flags":1,"s2_open_access_pdf_url":"http://arxiv.org/pdf/1311.3618","s2_open_access_landing_url":"https://www.semanticscholar.org/paper/18c125ce0f64e85577f7d30132cf0e92ec664bf4","s2_open_access_license":null,"s2_open_access_status":"GREEN","pmc_open_access_pdf_url":null,"pmc_open_access_landing_url":null,"pmc_open_access_license":null,"pmc_open_access_status":null,"unpaywall_open_access_pdf_url":null,"unpaywall_open_access_landing_url":null,"unpaywall_open_access_license":null,"unpaywall_open_access_status":null,"abstract":"Patterns and textures are key characteristics of many natural objects: a shirt can be striped, the wings of a butterfly can be veined, and the skin of an animal can be scaly. Aiming at supporting this dimension in image understanding, we address the problem of describing textures with semantic attributes. We identify a vocabulary of forty-seven texture terms and use them to describe a large dataset of patterns collected \"in the wild\". The resulting Describable Textures Dataset (DTD) is a basis to seek the best representation for recognizing describable texture attributes in images. We port from object recognition to texture recognition the Improved Fisher Vector (IFV) and Deep Convolutional-network Activation Features (DeCAF), and show that surprisingly, they both outperform specialized texture descriptors not only on our problem, but also in established material recognition datasets. We also show that our describable attributes are excellent texture descriptors, transferring between datasets and tasks, in particular, combined with IFV and DeCAF, they significantly outperform the state-of-the-art by more than 10% on both FMD and KTH-TIPS-2b benchmarks. We also demonstrate that they produce intuitive descriptions of materials and Internet images.","claims":[{"public_id":"cl_9939bf493e3767b71c59dec109236133","status":"active","text":"A vocabulary of forty-seven texture terms was identified to describe a large in-the-wild texture dataset.","confidence":0.98,"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/claims/cl_9939bf493e3767b71c59dec109236133"},{"public_id":"cl_2f3b53e1134dc824ca2df76a8ae5deac","status":"active","text":"Describable attributes produce intuitive descriptions of materials and Internet images.","confidence":0.92,"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/claims/cl_2f3b53e1134dc824ca2df76a8ae5deac"},{"public_id":"cl_6580f3b2a31cb68fedd43ff349d87c86","status":"active","text":"Describable attributes transfer well between datasets and tasks and, when combined with Improved Fisher Vector or Deep Convolutional-network Activation Features, improve the state of the art by more than 10% on the FMD and KTH-TIPS-2b benchmarks.","confidence":0.98,"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/claims/cl_6580f3b2a31cb68fedd43ff349d87c86"},{"public_id":"cl_17502f6a8e10a7495a7c666dba314424","status":"active","text":"Improved Fisher Vector and Deep Convolutional-network Activation Features outperform specialized texture descriptors on texture recognition and also on established material recognition datasets.","confidence":0.96,"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/claims/cl_17502f6a8e10a7495a7c666dba314424"}],"concepts":[{"public_id":"co_06f89f118c59d534141eea7b90b5b47f","status":"active","name":"FMD","description":"A benchmark dataset for material recognition.","types":["benchmark","dataset"],"aliases":["Flickr Material Database"],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_06f89f118c59d534141eea7b90b5b47f"},{"public_id":"co_3f04025bb46b08450c57b16d307db0b9","status":"active","name":"material recognition datasets","description":"Benchmark datasets used to evaluate recognition of material categories or properties.","types":["dataset","benchmark"],"aliases":[],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_3f04025bb46b08450c57b16d307db0b9"},{"public_id":"co_5cd08c3f4a1d774fb063ce42759f1809","status":"active","name":"Deep Convolutional-network Activation Features","description":"Feature representations derived from activations of deep convolutional networks and used here for image and texture recognition.","types":["feature representation","method"],"aliases":["DeCAF"],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_5cd08c3f4a1d774fb063ce42759f1809"},{"public_id":"co_6a3ef14095d9ac7b4247bff8cf23d63f","status":"active","name":"Improved Fisher Vector","description":"A feature representation originally used for object recognition and applied here to texture recognition.","types":["feature representation","method"],"aliases":["IFV"],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_6a3ef14095d9ac7b4247bff8cf23d63f"},{"public_id":"co_6b99c0710cc527867db0628b5b300915","status":"active","name":"materials","description":"Real-world surface or substance categories depicted in images.","types":["image content","object category"],"aliases":[],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_6b99c0710cc527867db0628b5b300915"},{"public_id":"co_761fc44d23e812742dc21d468af7e9c4","status":"active","name":"Describable Textures Dataset","description":"A large collection of texture and pattern images gathered in the wild and annotated with describable texture attributes.","types":["dataset"],"aliases":["DTD"],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_761fc44d23e812742dc21d468af7e9c4"},{"public_id":"co_931c93877ed80a54718d47a586524ead","status":"active","name":"Internet images","description":"Images collected from the web and used here for qualitative description examples.","types":["image collection"],"aliases":[],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_931c93877ed80a54718d47a586524ead"},{"public_id":"co_9a6591962b1fd14db2e094c92d2dc568","status":"active","name":"KTH-TIPS-2b","description":"A benchmark dataset for material recognition under varying imaging conditions.","types":["benchmark","dataset"],"aliases":[],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_9a6591962b1fd14db2e094c92d2dc568"},{"public_id":"co_b8c21aa37996ab11ac7c71cebf53798e","status":"active","name":"describable attributes","description":"Semantic texture attributes used to characterize patterns and textures in images.","types":["attributes","representation"],"aliases":["describable texture attributes"],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_b8c21aa37996ab11ac7c71cebf53798e"},{"public_id":"co_bb078cc57b4778f3c0e776cdbcfca622","status":"active","name":"specialized texture descriptors","description":"Handcrafted descriptors designed specifically for characterizing texture images.","types":["baseline","feature descriptor"],"aliases":[],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_bb078cc57b4778f3c0e776cdbcfca622"},{"public_id":"co_ce275986d2a136202b8f11aedfaa433c","status":"active","name":"texture recognition","description":"The task of identifying texture categories or properties from images.","types":["task"],"aliases":[],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_ce275986d2a136202b8f11aedfaa433c"},{"public_id":"co_eb24972f15fe77e85b155c3476f145b1","status":"active","name":"texture terms","description":"A vocabulary of forty-seven semantic words used to describe textures in images.","types":["vocabulary","semantic attributes"],"aliases":["texture vocabulary"],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_eb24972f15fe77e85b155c3476f145b1"}],"external_ids":{"DOI":"10.1109/CVPR.2014.461","ArXiv":"1311.3618","PubMed":null,"PubMedCentral":null,"MAG":2047643928,"DBLP":"conf/cvpr/CimpoiMKMV14","ACL":null},"open_access":{"is_open_access":true,"pdf_url":"http://arxiv.org/pdf/1311.3618","landing_url":"https://www.semanticscholar.org/paper/18c125ce0f64e85577f7d30132cf0e92ec664bf4","source":"semantic_scholar","pdf_url_source":"semantic_scholar_open_access_pdf","license":null,"status":"GREEN","reason":null},"reference_availability":{"status":"available","references_indexed":true,"full_text_available":true,"full_text_source":"arxiv","count_basis":"semantic_scholar_metadata","extraction_status":"not_applicable","reason":null},"source":{"provider":"episteme2","base_corpus":"semantic_scholar_dump","freshness_mode":"unknown","basis":["semantic_scholar_metadata","postgres_metadata"],"limits":["paper metadata is based on indexed upstream scholarly datasets","claims and concepts are available only for extracted papers","absence of claims or concepts means no extracted graph data is available in this response"],"status":"available","degraded":false,"degraded_reasons":[],"diagnostics":{"status":"available","degraded":false,"degraded_reasons":[],"metadata_status":"available","graph_status":"available","abstract_status":"available"},"source_flags":1},"paper_id":637485,"paper_uid":"af0bedc5-91f8-4602-87fe-1ff6e91ecd09","canonical_identity":{"paper_id":637485,"paper_uid":"af0bedc5-91f8-4602-87fe-1ff6e91ecd09","identity_status":"available","lookup_basis":"semantic_scholar_external_id","compatibility_path":"corpus_id"},"url":"https://sah.borca.ai/papers/4309276"}