{"id":72,"date":"2019-12-09T16:56:36","date_gmt":"2019-12-09T15:56:36","guid":{"rendered":"https:\/\/project.inria.fr\/desed\/?page_id=72"},"modified":"2020-04-20T15:19:36","modified_gmt":"2020-04-20T13:19:36","slug":"download","status":"publish","type":"page","link":"https:\/\/project.inria.fr\/desed\/download\/","title":{"rendered":"Download"},"content":{"rendered":"<p>The dataset is composed of two subsets:<\/p>\n<ul>\n<li><a href=\"https:\/\/project.inria.fr\/desed\/real-data\/\">Recorded soundscapes<\/a><\/li>\n<li><a href=\"https:\/\/project.inria.fr\/desed\/synthetic-data\/\">Synthetic soundscapes<\/a><\/li>\n<\/ul>\n<p>Links to the zenodo repos: <strong><a href=\"https:\/\/doi.org\/10.5281\/zenodo.3550598\" rel=\"nofollow\">DESED_synthetic<\/a><\/strong>, <strong><a href=\"https:\/\/zenodo.org\/record\/3565749\" rel=\"nofollow\">DESED_real<\/a><\/strong><\/p>\n<p>&nbsp;<\/p>\n<p><strong>After downloading the data (see below) you should have this tree:<\/strong><\/p>\n<pre><code>\u251c\u2500\u2500 dcase2019\r\n\u2502\u00a0\u00a0 \u251c\u2500\u2500 dataset\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u251c\u2500\u2500 audio\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u251c\u2500\u2500 eval\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u251c\u2500\u2500 500ms\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u251c\u2500\u2500 5500ms\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u251c\u2500\u2500 9500ms\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u251c\u2500\u2500 distorted_clipping\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u251c\u2500\u2500 distorted_drc\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u251c\u2500\u2500 distorted_highpass_filter\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u251c\u2500\u2500 distorted_lowpass_filter\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u251c\u2500\u2500 distorted_smartphone_playback\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u251c\u2500\u2500 distorted_smartphone_recording\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u251c\u2500\u2500 fbsnr_0dB\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u251c\u2500\u2500 fbsnr_15dB\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u251c\u2500\u2500 fbsnr_24dB\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u251c\u2500\u2500 fbsnr_30dB\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u251c\u2500\u2500 ls_0dB\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u251c\u2500\u2500 ls_15dB\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2514\u2500\u2500 ls_30dB\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u251c\u2500\u2500 train\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u251c\u2500\u2500 synthetic\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u251c\u2500\u2500 unlabel_in_domain\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2514\u2500\u2500 weak\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2514\u2500\u2500 validation\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2514\u2500\u2500 metadata\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0     \u251c\u2500\u2500 eval\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0     \u251c\u2500\u2500 train\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0     \u2514\u2500\u2500 validation\r\n\u2502\u00a0\u00a0 \u2514\u2500\u2500 src\r\n\u251c\u2500\u2500 real_data                                   (subpart of dcase2019)\r\n\u2502\u00a0\u00a0 \u251c\u2500\u2500 audio\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u251c\u2500\u2500 train\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502   \u251c\u2500\u2500 unlabel_in_domain\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502   \u2514\u2500\u2500 weak\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2514\u2500\u2500 validation\r\n\u2502\u00a0\u00a0 \u251c\u2500\u2500 metadata\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u251c\u2500\u2500 train\r\n\u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2514\u2500\u2500 validation\r\n\u2502\u00a0\u00a0 \u2514\u2500\u2500 src\r\n\u2514\u2500\u2500 synthetic\r\n    \u251c\u2500\u2500 audio\r\n    \u2502\u00a0\u00a0 \u251c\u2500\u2500 eval\r\n    \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u251c\u2500\u2500 distorted_fbsnr_30dB            (6 subfolders for each distortion, audio are directly given because a matlab code has been used to generate them) \r\n    \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2514\u2500\u2500 soundbank                       (Raw (bank of) data that can be used to create synthetic data)\r\n    \u2502\u00a0\u00a0 \u2502\u00a0\u00a0     \u251c\u2500\u2500 background                  (2 subfolders, youtube and freesound)\r\n    \u2502\u00a0\u00a0 \u2502\u00a0\u00a0     \u251c\u2500\u2500 background_long             (5 subfolders)\r\n    \u2502\u00a0\u00a0 \u2502\u00a0\u00a0     \u251c\u2500\u2500 foreground                  (18 subfolders)\r\n    \u2502\u00a0\u00a0 \u2502\u00a0\u00a0     \u251c\u2500\u2500 foreground_on_off           (10 subfolders)\r\n    \u2502\u00a0\u00a0 \u2502\u00a0\u00a0     \u2514\u2500\u2500 foreground_short            (5 subfolders)\r\n    \u2502\u00a0\u00a0 \u2514\u2500\u2500 train\r\n    \u2502\u00a0\u00a0     \u251c\u2500\u2500 soundbank                       (Raw (bank of) data that can be used to create synthetic data)\r\n    \u2502\u00a0\u00a0     \u2502\u00a0\u00a0 \u251c\u2500\u2500 background\r\n    \u2502\u00a0\u00a0     \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2514\u2500\u2500 sins                    (Has to be downloaded by: get_background_training.py)\r\n    \u2502\u00a0\u00a0     \u2502\u00a0\u00a0 \u2514\u2500\u2500 foreground                  (14 subfolders)\r\n    \u2502\u00a0\u00a0     \u2514\u2500\u2500 synthetic\r\n    \u251c\u2500\u2500 metadata\r\n    \u2502\u00a0\u00a0 \u251c\u2500\u2500 eval\r\n    \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2514\u2500\u2500 soundscapes                     (metadata to reproduce the wav files used in dcase2019)\r\n    \u2502\u00a0\u00a0 \u2502\u00a0\u00a0     \u251c\u2500\u2500 500ms\r\n    \u2502\u00a0\u00a0 \u2502\u00a0\u00a0     \u251c\u2500\u2500 5500ms\r\n    \u2502\u00a0\u00a0 \u2502\u00a0\u00a0     \u251c\u2500\u2500 9500ms\r\n    \u2502\u00a0\u00a0 \u2502\u00a0\u00a0     \u251c\u2500\u2500 fbsnr_0dB\r\n    \u2502\u00a0\u00a0 \u2502\u00a0\u00a0     \u251c\u2500\u2500 fbsnr_15dB\r\n    \u2502\u00a0\u00a0 \u2502\u00a0\u00a0     \u251c\u2500\u2500 fbsnr_24dB\r\n    \u2502\u00a0\u00a0 \u2502\u00a0\u00a0     \u251c\u2500\u2500 fbsnr_30dB\r\n    \u2502\u00a0\u00a0 \u2502\u00a0\u00a0     \u251c\u2500\u2500 ls_0dB\r\n    \u2502\u00a0\u00a0 \u2502\u00a0\u00a0     \u251c\u2500\u2500 ls_15dB\r\n    \u2502\u00a0\u00a0 \u2502\u00a0\u00a0     \u2514\u2500\u2500 ls_30dB\r\n    \u2502\u00a0\u00a0 \u2514\u2500\u2500 train\r\n    \u2502\u00a0\u00a0     \u2514\u2500\u2500 soundscapes                     (metadata to reproduce the wav files used in dcase2019)\r\n    \u2514\u2500\u2500 src                                     (Source code to regenerate the dcase2019 dataset or generate new mixtures)\r\n<\/code><\/pre>\n","protected":false},"excerpt":{"rendered":"<p>The dataset is composed of two subsets: Recorded soundscapes Synthetic soundscapes Links to the zenodo repos: DESED_synthetic, DESED_real &nbsp; After downloading the data (see below) you should have this tree: \u251c\u2500\u2500 dcase2019 \u2502\u00a0\u00a0 \u251c\u2500\u2500 dataset \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u251c\u2500\u2500 audio \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u251c\u2500\u2500 eval \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u2502\u00a0\u00a0 \u251c\u2500\u2500 500ms\u2026<\/p>\n<p> <a class=\"continue-reading-link\" href=\"https:\/\/project.inria.fr\/desed\/download\/\"><span>Continue reading<\/span><i class=\"crycon-right-dir\"><\/i><\/a> <\/p>\n","protected":false},"author":1380,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"footnotes":""},"class_list":["post-72","page","type-page","status-publish","hentry"],"_links":{"self":[{"href":"https:\/\/project.inria.fr\/desed\/wp-json\/wp\/v2\/pages\/72","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/project.inria.fr\/desed\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/project.inria.fr\/desed\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/project.inria.fr\/desed\/wp-json\/wp\/v2\/users\/1380"}],"replies":[{"embeddable":true,"href":"https:\/\/project.inria.fr\/desed\/wp-json\/wp\/v2\/comments?post=72"}],"version-history":[{"count":11,"href":"https:\/\/project.inria.fr\/desed\/wp-json\/wp\/v2\/pages\/72\/revisions"}],"predecessor-version":[{"id":216,"href":"https:\/\/project.inria.fr\/desed\/wp-json\/wp\/v2\/pages\/72\/revisions\/216"}],"wp:attachment":[{"href":"https:\/\/project.inria.fr\/desed\/wp-json\/wp\/v2\/media?parent=72"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}