{"id":39714,"date":"2026-03-17T07:18:20","date_gmt":"2026-03-17T06:18:20","guid":{"rendered":"https:\/\/www.graviton.at\/letterswaplibrary\/anyone-has-any-good-rir-mega-dataset-in-the-audio-ml-space-synthetic\/"},"modified":"2026-03-17T07:18:20","modified_gmt":"2026-03-17T06:18:20","slug":"anyone-has-any-good-rir-mega-dataset-in-the-audio-ml-space-synthetic","status":"publish","type":"post","link":"https:\/\/www.graviton.at\/letterswaplibrary\/anyone-has-any-good-rir-mega-dataset-in-the-audio-ml-space-synthetic\/","title":{"rendered":"Anyone Has Any Good RIR Mega Dataset In The Audio ML Space? [Synthetic]"},"content":{"rendered":"<p><!-- SC_OFF --><\/p>\n<div class=\"md\">\n<p>Came across this dataset paper that I think deserves more attention. <\/p>\n<p>RIR-Mega is a large-scale collection of simulated Room Impulse Responses (RIRs) designed specifically for ML workflows. What makes it stand out from older RIR datasets:<\/p>\n<ul>\n<li>50,000 RIRs with a clean, flat Parquet metadata schema (RT60, DRR, C50, C80, band RT60s)<\/li>\n<li>Three evaluation splits: random, unseen_room, and unseen_distance \u2014 so you can actually test generalization<\/li>\n<\/ul>\n<p>The HF dataset is at: <a href=\"https:\/\/huggingface.co\/datasets\/mandipgoswami\/rirmega\">https:\/\/huggingface.co\/datasets\/mandipgoswami\/rirmega<\/a> Paper: <a href=\"https:\/\/arxiv.org\/abs\/2510.18917\">https:\/\/arxiv.org\/abs\/2510.18917<\/a><\/p>\n<p>Has anyone used this for dereverberation or acoustic parameter estimation? Curious how it holds up against BUT-ReverbDB or OpenRIR for downstream ASR robustness tasks.<\/p>\n<\/div>\n<p><!-- SC_ON -->   submitted by   <a href=\"https:\/\/www.reddit.com\/user\/Stellar_Bluebird\"> \/u\/Stellar_Bluebird <\/a> <br \/> <span><a href=\"https:\/\/www.reddit.com\/r\/datasets\/comments\/1rvyevx\/anyone_has_any_good_rir_mega_dataset_in_the_audio\/\">[link]<\/a><\/span>   <span><a href=\"https:\/\/www.reddit.com\/r\/datasets\/comments\/1rvyevx\/anyone_has_any_good_rir_mega_dataset_in_the_audio\/\">[comments]<\/a><\/span><\/p><div class='watch-action'><div class='watch-position align-right'><div class='action-like'><a class='lbg-style1 like-39714 jlk' href='javascript:void(0)' data-task='like' data-post_id='39714' data-nonce='bc39e8310e' 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-39714 lc'>0<\/span><\/a><\/div><\/div> <div class='status-39714 status align-right'><\/div><\/div><div class='wti-clear'><\/div>","protected":false},"excerpt":{"rendered":"<p>Came across this dataset paper that I think deserves more attention. RIR-Mega is a large-scale collection of&#8230;<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[85],"tags":[],"class_list":["post-39714","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\/39714","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"}],"author":[{"embeddable":true,"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/comments?post=39714"}],"version-history":[{"count":0,"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/posts\/39714\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/media?parent=39714"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/categories?post=39714"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/tags?post=39714"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}