{ "access": { "embargo": { "active": false, "reason": null }, "files": "public", "record": "public", "status": "open" }, "created": "2017-12-07T22:23:52.411765+00:00", "custom_fields": { "thesis:university": "Universitat Pompeu Fabra, Barcelona" }, "deletion_status": { "is_deleted": false, "status": "P" }, "files": { "count": 1, "enabled": true, "entries": { "ErruzG_mthesis.pdf": { "checksum": "md5:d9744762d1ae9d99afda967749046ff9", "ext": "pdf", "id": "f977f6d6-d93a-46a6-9936-88a698e502b0", "key": "ErruzG_mthesis.pdf", "metadata": null, "mimetype": "application/pdf", "size": 4617089 } }, "order": [], "total_bytes": 4617089 }, "id": "1095835", "is_draft": false, "is_published": true, "links": { "access": "https://zenodo.org/api/records/1095835/access", "access_links": "https://zenodo.org/api/records/1095835/access/links", "access_request": "https://zenodo.org/api/records/1095835/access/request", "access_users": "https://zenodo.org/api/records/1095835/access/users", "archive": "https://zenodo.org/api/records/1095835/files-archive", "archive_media": "https://zenodo.org/api/records/1095835/media-files-archive", "communities": "https://zenodo.org/api/records/1095835/communities", "communities-suggestions": "https://zenodo.org/api/records/1095835/communities-suggestions", "doi": "https://doi.org/10.5281/zenodo.1095835", "draft": "https://zenodo.org/api/records/1095835/draft", "files": "https://zenodo.org/api/records/1095835/files", "latest": "https://zenodo.org/api/records/1095835/versions/latest", "latest_html": "https://zenodo.org/records/1095835/latest", "media_files": "https://zenodo.org/api/records/1095835/media-files", "parent": "https://zenodo.org/api/records/1095834", "parent_doi": "https://zenodo.org/doi/10.5281/zenodo.1095834", "parent_html": "https://zenodo.org/records/1095834", "requests": "https://zenodo.org/api/records/1095835/requests", "reserve_doi": "https://zenodo.org/api/records/1095835/draft/pids/doi", "self": "https://zenodo.org/api/records/1095835", "self_doi": "https://zenodo.org/doi/10.5281/zenodo.1095835", "self_html": "https://zenodo.org/records/1095835", "self_iiif_manifest": "https://zenodo.org/api/iiif/record:1095835/manifest", "self_iiif_sequence": "https://zenodo.org/api/iiif/record:1095835/sequence/default", "versions": "https://zenodo.org/api/records/1095835/versions" }, "media_files": { "count": 0, "enabled": false, "entries": {}, "order": [], "total_bytes": 0 }, "metadata": { "contributors": [ { "affiliations": [ { "name": "Universitat Pompeu Fabra, Barcelona" } ], "person_or_org": { "family_name": "Miron", "given_name": "Marius", "name": "Miron, Marius", "type": "personal" }, "role": { "id": "supervisor", "title": { "de": "SupervisorIn", "en": "Supervisor" } } }, { "affiliations": [ { "name": "Eurecat" } ], "person_or_org": { "family_name": "Garriga", "given_name": "Adan", "name": "Garriga, Adan", "type": "personal" }, "role": { "id": "supervisor", "title": { "de": "SupervisorIn", "en": "Supervisor" } } } ], "creators": [ { "affiliations": [ { "name": "Universitat Pompeu Fabra, Barcelona" } ], "person_or_org": { "family_name": "Erruz", "given_name": "Gerard", "name": "Erruz, Gerard", "type": "personal" } } ], "description": "
This work is a study on source separation techniques for binaural music mixtures. The chosen framework uses a Convolutional Neural Network (CNN) to estimate time-frequency soft masks. This masks are used to extract the different sources from the original two-channel mixture signal. Its baseline single-channel architecture performed state-of-the-art results on monaural music mixtures under low-latency conditions. It has been extended to perform separation in two-channel signals, being the first two-channel CNN joint estimation architecture. This means that filters are learned for each source by taking in account both channels information. Furthermore, a specific binaural condition is included during training stage. It uses Interaural Level Difference (ILD) information to improve spatial images of extracted sources. Concurrently, we present a novel tool to create binaural scenes for testing purposes. Multiple binaural scenes are rendered from a music dataset of four instruments (voice, drums, bass and others). The CNN framework have been tested for these binaural scenes and compared with monaural and stereo results. The system showed a great amount of adaptability and good separation results in all the scenarios. These results are used to evaluate spatial information impact on separation performance.
", "languages": [ { "id": "eng", "title": { "en": "English" } } ], "publication_date": "2017-12-07", "publisher": "Zenodo", "related_identifiers": [ { "identifier": "10.5281/zenodo.884105", "relation_type": { "id": "documents", "title": { "de": "Dokumentiert", "en": "Documents" } }, "scheme": "doi" } ], "resource_type": { "id": "publication-other", "title": { "de": "Sonstige", "en": "Other" } }, "rights": [ { "description": { "en": "The Creative Commons Attribution license allows re-distribution and re-use of a licensed work on the condition that the creator is appropriately credited." }, "icon": "cc-by-icon", "id": "cc-by-4.0", "props": { "scheme": "spdx", "url": "https://creativecommons.org/licenses/by/4.0/legalcode" }, "title": { "en": "Creative Commons Attribution 4.0 International" } } ], "subjects": [ { "subject": "Audio Source Separation" }, { "subject": "Neural Networks" }, { "subject": "Music Source Separation" }, { "subject": "Binaural" } ], "title": "Binaural Source Separation with Convolutional Neural Networks" }, "parent": { "access": { "owned_by": { "user": 35601 } }, "communities": { "default": "78c30cbc-1c1b-4ac2-a20b-7715387167e0", "entries": [ { "access": { "member_policy": "open", "members_visibility": "public", "record_policy": "open", "review_policy": "open", "visibility": "public" }, "children": { "allow": false }, "created": "2017-11-22T11:21:43.788681+00:00", "custom_fields": {}, "deletion_status": { "is_deleted": false, "status": "P" }, "id": "78c30cbc-1c1b-4ac2-a20b-7715387167e0", "links": {}, "metadata": { "curation_policy": "Master thesis presented and approved by the Master in Sound and Music Computing of the Universitat Pompeu Fabra, Barcelona.
\r\n", "description": "Thesis of the Master in Sound and Music Computing (https://www.upf.edu/web/smc) of the Universitat Pompeu Fabra, Barcelona (https://www.upf.edu)", "page": "", "title": "Master in Sound and Music Computing" }, "revision_id": 0, "slug": "smc-master", "updated": "2020-04-28T17:00:02.532998+00:00" } ], "ids": [ "78c30cbc-1c1b-4ac2-a20b-7715387167e0" ] }, "id": "1095834", "pids": { "doi": { "client": "datacite", "identifier": "10.5281/zenodo.1095834", "provider": "datacite" } } }, "pids": { "doi": { "client": "datacite", "identifier": "10.5281/zenodo.1095835", "provider": "datacite" }, "oai": { "identifier": "oai:zenodo.org:1095835", "provider": "oai" } }, "revision_id": 9, "stats": { "all_versions": { "data_volume": 3342772436.0, "downloads": 724, "unique_downloads": 675, "unique_views": 351, "views": 364 }, "this_version": { "data_volume": 3342772436.0, "downloads": 724, "unique_downloads": 675, "unique_views": 350, "views": 363 } }, "status": "published", "updated": "2020-01-20T17:41:58.607310+00:00", "versions": { "index": 1, "is_latest": true } }