{"corpus_id":6258614,"paper_sha":"7002d8c61be9f1ea210f88059df6955c88db62b7","doi":"10.1109/CVPR.2018.00016","arxiv_id":"1711.08565","pmid":null,"pmcid":null,"mag_id":2769994766,"dblp_id":"journals/corr/abs-1711-08565","acl_id":null,"title":"Person Transfer GAN to Bridge Domain Gap for Person Re-identification","year":2017,"publication_date":"2017-11-23","venue":"2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition","journal":{"name":"2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition","pages":"79-88","volume":null},"journal_issn":null,"journal_title":null,"publication_types":["JournalArticle","Conference"],"pubmed_pub_types":null,"s2_fields_of_study":["Computer Science"],"reference_count":45,"citation_count":1908,"influential_citation_count":413,"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":"https://arxiv.org/pdf/1711.08565","s2_open_access_landing_url":"https://www.semanticscholar.org/paper/7002d8c61be9f1ea210f88059df6955c88db62b7","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":"Although the performance of person Re-Identification (ReID) has been significantly boosted, many challenging issues in real scenarios have not been fully investigated, e.g., the complex scenes and lighting variations, viewpoint and pose changes, and the large number of identities in a camera network. To facilitate the research towards conquering those issues, this paper contributes a new dataset called MSMT171 with many important features, e.g., 1) the raw videos are taken by an 15-camera network deployed in both indoor and outdoor scenes, 2) the videos cover a long period of time and present complex lighting variations, and 3) it contains currently the largest number of annotated identities, i.e., 4,101 identities and 126,441 bounding boxes. We also observe that, domain gap commonly exists between datasets, which essentially causes severe performance drop when training and testing on different datasets. This results in that available training data cannot be effectively leveraged for new testing domains. To relieve the expensive costs of annotating new training samples, we propose a Person Transfer Generative Adversarial Network (PTGAN) to bridge the domain gap. Comprehensive experiments show that the domain gap could be substantially narrowed-down by the PTGAN.","claims":[{"public_id":"cl_1f3b9f979e8e8521f2bf96d9ef78555e","status":"active","text":"Comprehensive experiments show that Person Transfer Generative Adversarial Network substantially narrows the domain gap.","confidence":0.95,"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_1f3b9f979e8e8521f2bf96d9ef78555e"},{"public_id":"cl_7db947deae66604cfbf4f405181bc884","status":"active","text":"Domain gap between person re-identification datasets causes severe performance drops when models are trained and tested on different datasets.","confidence":0.93,"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_7db947deae66604cfbf4f405181bc884"},{"public_id":"cl_cf4ea864e85a56f94b1e48970c3c249c","status":"active","text":"MSMT171 is a new person re-identification dataset with 15-camera indoor and outdoor videos, long-term capture, complex lighting variations, 4,101 identities, and 126,441 bounding boxes.","confidence":0.99,"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_cf4ea864e85a56f94b1e48970c3c249c"},{"public_id":"cl_7eb3513ecbf24e2d9947e1b669482d92","status":"active","text":"Person Transfer Generative Adversarial Network is proposed to bridge the dataset domain gap and reduce the need for expensive annotation of new training samples.","confidence":0.97,"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_7eb3513ecbf24e2d9947e1b669482d92"}],"concepts":[{"public_id":"co_2eb0faedc8fe0d8638d6458c7bdf0dcf","status":"active","name":"15-camera network","description":"A camera network of 15 cameras used to collect the MSMT171 videos.","types":["data collection setup"],"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_2eb0faedc8fe0d8638d6458c7bdf0dcf"},{"public_id":"co_326853a138c90963d2bb1c425f752836","status":"active","name":"annotation of new training samples","description":"The manual labeling of additional training data for a new target domain.","types":["process"],"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_326853a138c90963d2bb1c425f752836"},{"public_id":"co_48ae7ba98508f3ccd58202f2cd0ef7c3","status":"active","name":"performance drop","description":"A decrease in model accuracy or effectiveness when evaluated on a different dataset or domain.","types":["outcome"],"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_48ae7ba98508f3ccd58202f2cd0ef7c3"},{"public_id":"co_5bdba6929687afe5a717ebb726867a78","status":"active","name":"person re-identification","description":"The task of matching person identities across images or video frames from different cameras.","types":["task"],"aliases":["ReID"],"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_5bdba6929687afe5a717ebb726867a78"},{"public_id":"co_6e9ec09c26c527cb3cf579b9122c92f1","status":"active","name":"domain gap","description":"The mismatch between source and target datasets that affects model transfer across domains.","types":["problem"],"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_6e9ec09c26c527cb3cf579b9122c92f1"},{"public_id":"co_81f6d1654dffecb3ad90ea427cd58455","status":"active","name":"person re-identification dataset","description":"A dataset used for identifying people across different camera views.","types":["dataset"],"aliases":["ReID dataset"],"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_81f6d1654dffecb3ad90ea427cd58455"},{"public_id":"co_82f7c2927fa3b803e9fded670d41f9bc","status":"active","name":"comprehensive experiments","description":"An experimental evaluation covering multiple tests of the proposed method.","types":["evaluation"],"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_82f7c2927fa3b803e9fded670d41f9bc"},{"public_id":"co_da26942f779113266ed2e99c8b564f59","status":"active","name":"Person Transfer Generative Adversarial Network","description":"A generative adversarial network designed to transfer person images across datasets to reduce cross-domain differences.","types":["method"],"aliases":["PTGAN"],"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_da26942f779113266ed2e99c8b564f59"},{"public_id":"co_da8d6541ddff0d7c00d23665b32608c0","status":"active","name":"MSMT171","description":"A person re-identification dataset collected from a 15-camera network with indoor and outdoor scenes, long-term recordings, and many annotated identities.","types":["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_da8d6541ddff0d7c00d23665b32608c0"}],"external_ids":{"DOI":"10.1109/CVPR.2018.00016","ArXiv":"1711.08565","PubMed":null,"PubMedCentral":null,"MAG":2769994766,"DBLP":"journals/corr/abs-1711-08565","ACL":null},"open_access":{"is_open_access":true,"pdf_url":"https://arxiv.org/pdf/1711.08565","landing_url":"https://www.semanticscholar.org/paper/7002d8c61be9f1ea210f88059df6955c88db62b7","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":636029,"paper_uid":"84c33559-240d-4a0a-9b0c-e1f468d2c084","canonical_identity":{"paper_id":636029,"paper_uid":"84c33559-240d-4a0a-9b0c-e1f468d2c084","identity_status":"available","lookup_basis":"semantic_scholar_external_id","compatibility_path":"corpus_id"},"url":"https://sah.borca.ai/papers/6258614"}