{"corpus_id":4322589,"paper_sha":"cceec617594f0238a384903fabcac0c0ca1b2b35","doi":null,"arxiv_id":"1704.07711","pmid":null,"pmcid":null,"mag_id":2608346182,"dblp_id":"journals/corr/MinaeeW17a","acl_id":null,"title":"Masked Signal Decomposition Using Subspace Representation and Its Applications","year":2017,"publication_date":"2017-04-25","venue":"arXiv.org","journal":{"name":"ArXiv","pages":null,"volume":"abs/1704.07711"},"journal_issn":null,"journal_title":null,"publication_types":["JournalArticle"],"pubmed_pub_types":null,"s2_fields_of_study":["Computer Science"],"reference_count":41,"citation_count":12,"influential_citation_count":0,"is_open_access":false,"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":null,"s2_open_access_landing_url":null,"s2_open_access_license":null,"s2_open_access_status":null,"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":"Signal decomposition is a classical problem in signal processing, which aims to separate an observed signal into two or more components each with its own property. Usually each component is described by its own subspace or dictionary. Extensive research has been done for the case where the components are additive, but in real world applications, the components are often non-additive. For example, an image may consist of a foreground object overlaid on a background, where each pixel either belongs to the foreground or the background. In such a situation, to separate signal components, we need to find a binary mask which shows the location of each component. Therefore it requires to solve a binary optimization problem. Since most of the binary optimization problems are intractable, we relax this problem to the approximated continuous problem, and solve it by alternating optimization technique, which solves the mask and the subspace projection for each component, alternatingly and iteratively. We show the application of the proposed algorithm for three applications: separation of text from background in images, separation of moving objects from a background undergoing global camera motion in videos, separation of sinusoidal and spike components in one dimensional signals. We demonstrate in each case that considering the non-additive nature of the problem can lead to significant improvement.","claims":[{"public_id":"cl_872e5c155acf634228a4196612aa929c","status":"active","text":"Considering the non-additive nature of the problem yields significant improvement in text-background separation, moving-object separation under global camera motion, and separation of sinusoidal and spike components.","confidence":0.9,"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_872e5c155acf634228a4196612aa929c"},{"public_id":"cl_a9c93dcc1f820831dfc45869f7cd440d","status":"active","text":"Non-additive signal components can be separated by introducing a binary mask that indicates the location of each component.","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_a9c93dcc1f820831dfc45869f7cd440d"},{"public_id":"cl_c1f835cc4e96e118a19688f3205982b9","status":"active","text":"The resulting binary optimization problem is relaxed to a continuous approximation and solved with alternating optimization over the mask and the subspace projection for each component.","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_c1f835cc4e96e118a19688f3205982b9"}],"concepts":[{"public_id":"co_0c1eeb2901ce9b9523669c3322f7ea49","status":"active","name":"non-additive signal components","description":"Signal components that combine through masking or overlap rather than by simple addition.","types":["phenomenon"],"aliases":["non-additive components"],"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_0c1eeb2901ce9b9523669c3322f7ea49"},{"public_id":"co_1038bb187544cc6ef7d6043ac3aeaef2","status":"active","name":"moving objects separation from background","description":"The video application of separating moving foreground objects from a background undergoing global camera motion.","types":["application"],"aliases":["moving object separation"],"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_1038bb187544cc6ef7d6043ac3aeaef2"},{"public_id":"co_299a7cc9608f89365f880e950607acb2","status":"active","name":"masked signal decomposition","description":"A signal decomposition approach that uses a mask to assign components to different regions of the observed signal.","types":["method"],"aliases":["mask-based signal decomposition"],"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_299a7cc9608f89365f880e950607acb2"},{"public_id":"co_2dc38b3f66c7c6b5219d6190a4355b87","status":"active","name":"sinusoidal and spike components","description":"A one-dimensional signal decomposition setting involving sinusoidal and impulsive spike parts.","types":["signal components"],"aliases":["sinusoidal components and spike components"],"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_2dc38b3f66c7c6b5219d6190a4355b87"},{"public_id":"co_3985beb3cd77ff048101b3f708e50a73","status":"active","name":"alternating optimization technique","description":"An iterative optimization procedure that alternates between solving for the mask and solving for the subspace projection.","types":["optimization technique"],"aliases":["alternating optimization"],"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_3985beb3cd77ff048101b3f708e50a73"},{"public_id":"co_4e09a0a5a173c4925a0f9dac422dfc67","status":"active","name":"continuous approximation","description":"A relaxed continuous version of the binary optimization problem that is easier to solve.","types":["optimization technique"],"aliases":["relaxation"],"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_4e09a0a5a173c4925a0f9dac422dfc67"},{"public_id":"co_53842a54f83644c1f3cf9f6861495741","status":"active","name":"binary mask","description":"A binary indicator field that marks which locations belong to each signal component.","types":["representation"],"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_53842a54f83644c1f3cf9f6861495741"},{"public_id":"co_573040d6104ef5ea3e4c9b954fd8e865","status":"active","name":"signal decomposition","description":"The task of separating an observed signal into multiple components with distinct properties.","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_573040d6104ef5ea3e4c9b954fd8e865"},{"public_id":"co_63f57060a4aba25e6de73c28917b0935","status":"active","name":"text separation from background","description":"The application of separating textual foreground from background content in images.","types":["application"],"aliases":["text from background separation"],"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_63f57060a4aba25e6de73c28917b0935"},{"public_id":"co_c344b906de53c638fd43ef8a06c61aad","status":"active","name":"subspace projection","description":"The projection of each component onto its corresponding subspace during optimization.","types":["operation"],"aliases":["projection"],"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_c344b906de53c638fd43ef8a06c61aad"},{"public_id":"co_c8f6a23a7ebed39a6b1f509e7cdaacf1","status":"active","name":"subspace representation","description":"A representation in which each signal component is modeled by its own subspace or dictionary.","types":["representation"],"aliases":["subspace","dictionary representation"],"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_c8f6a23a7ebed39a6b1f509e7cdaacf1"},{"public_id":"co_cfd8d4d603845a239d293bb2d1a8fd4e","status":"active","name":"binary optimization problem","description":"An optimization problem with binary decision variables used to estimate the component mask.","types":["optimization 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_cfd8d4d603845a239d293bb2d1a8fd4e"}],"external_ids":{"DOI":null,"ArXiv":"1704.07711","PubMed":null,"PubMedCentral":null,"MAG":2608346182,"DBLP":"journals/corr/MinaeeW17a","ACL":null},"open_access":{"is_open_access":true,"pdf_url":"https://arxiv.org/pdf/1704.07711","landing_url":"https://arxiv.org/abs/1704.07711","source":"arxiv","pdf_url_source":"derived_arxiv","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","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":633118,"paper_uid":"0a866935-1774-463c-af8b-c623d8694656","canonical_identity":{"paper_id":633118,"paper_uid":"0a866935-1774-463c-af8b-c623d8694656","identity_status":"available","lookup_basis":"semantic_scholar_external_id","compatibility_path":"corpus_id"},"url":"https://sah.borca.ai/papers/4322589"}