{"id":32440,"date":"2025-01-31T14:27:56","date_gmt":"2025-01-31T13:27:56","guid":{"rendered":"https:\/\/www.graviton.at\/letterswaplibrary\/help-creating-a-deepfake-audio-dataset\/"},"modified":"2025-01-31T14:27:56","modified_gmt":"2025-01-31T13:27:56","slug":"help-creating-a-deepfake-audio-dataset","status":"publish","type":"post","link":"https:\/\/www.graviton.at\/letterswaplibrary\/help-creating-a-deepfake-audio-dataset\/","title":{"rendered":"Help Creating A Deepfake Audio Dataset?"},"content":{"rendered":"<p><!-- SC_OFF --><\/p>\n<div class=\"md\">\n<p>Hey everyone,<\/p>\n<p>I\u2019m working on building a deepfake audio dataset and wanted to get some help on best practices. I want to ensure that the dataset is diverse and representative for training an effective detection model.<\/p>\n<p>Some questions I have:<\/p>\n<p>How many speakers should I aim for to get a balanced dataset?<\/p>\n<p>Should I maintain an equal gender ratio, or does it make a difference ?<\/p>\n<p>How long is enough from each source(mins, hours)<\/p>\n<p>Any recommended sources or strategies for collecting high-quality real audio?<\/p>\n<p>What sample rates (e.g., 16kHz, 44.1kHz, 48kHz) or a what mix?<\/p>\n<p>Are certain codecs (e.g., MP3, AAC, Opus, WAV) more challenging for detection models?<\/p>\n<p>Would love to hear from those who have experience <\/p>\n<\/div>\n<p><!-- SC_ON -->   submitted by   <a href=\"https:\/\/www.reddit.com\/user\/Fuzzy_Cream_5073\"> \/u\/Fuzzy_Cream_5073 <\/a> <br \/> <span><a href=\"https:\/\/www.reddit.com\/r\/datasets\/comments\/1iedv21\/help_creating_a_deepfake_audio_dataset\/\">[link]<\/a><\/span>   <span><a href=\"https:\/\/www.reddit.com\/r\/datasets\/comments\/1iedv21\/help_creating_a_deepfake_audio_dataset\/\">[comments]<\/a><\/span><\/p><div class='watch-action'><div class='watch-position align-right'><div class='action-like'><a class='lbg-style1 like-32440 jlk' href='javascript:void(0)' data-task='like' data-post_id='32440' data-nonce='614a020375' rel='nofollow'><img class='wti-pixel' src='https:\/\/www.graviton.at\/letterswaplibrary\/wp-content\/plugins\/wti-like-post\/images\/pixel.gif' title='Like' \/><span class='lc-32440 lc'>0<\/span><\/a><\/div><\/div> <div class='status-32440 status align-right'><\/div><\/div><div class='wti-clear'><\/div>","protected":false},"excerpt":{"rendered":"<p>Hey everyone, I\u2019m working on building a deepfake audio dataset and wanted to get some help on&#8230;<\/p>\n","protected":false},"author":0,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[85],"tags":[],"class_list":["post-32440","post","type-post","status-publish","format-standard","hentry","category-datatards","wpcat-85-id"],"_links":{"self":[{"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/posts\/32440","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/types\/post"}],"replies":[{"embeddable":true,"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/comments?post=32440"}],"version-history":[{"count":0,"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/posts\/32440\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/media?parent=32440"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/categories?post=32440"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/tags?post=32440"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}