{"id":1635,"date":"2025-10-08T14:12:00","date_gmt":"2025-10-08T12:12:00","guid":{"rendered":"https:\/\/deepbirddetect.de\/?p=1635"},"modified":"2025-10-21T08:18:16","modified_gmt":"2025-10-21T06:18:16","slug":"neues-paper-bei-arxiv-2","status":"publish","type":"post","link":"https:\/\/deepbirddetect.de\/en\/neues-paper-bei-arxiv-2\/","title":{"rendered":"New Paper on ArXiv"},"content":{"rendered":"<p>Our paper <a href=\"https:\/\/arxiv.org\/abs\/2509.24901\" data-type=\"link\" data-id=\"https:\/\/arxiv.org\/abs\/2509.24901\">Unmute the Patch Tokens: Rethinking Probing in Multi-Label Audio Classification<\/a> by Lukas Rauch, Ren\u00e9 Heinrich, Houtan Ghaffari, Lukas Miklautz, Ilyass Moummad, Bernhard Sick and Christoph Scholz is online on\u00a0<em>ArXiv<\/em>\u00a0.<\/p>\n\n\n\n<p>Abstract: <br>Although probing frozen models has become a standard evaluation paradigm, self-supervised learning in audio defaults to fine-tuning. A key reason is that global pooling creates an information bottleneck causing linear probes to misrepresent the embedding quality: The\u00a0cls-token discards crucial token information about dispersed, localized events in multi-label audio. This weakness is rooted in the mismatch between the pretraining objective (operating globally) and the downstream task (localized events). Across a comprehensive benchmark of 13 datasets and 6 spectrogram-based encoders, we first investigate the global pooling bottleneck. We then introduce binarized prototypical probes: a lightweight and simple pooling method that learns prototypes to perform class-wise information aggregation. Despite its simplicity, our method notably outperforms linear and attentive probing. Our work establishes probing as a competitive and efficient paradigm for evaluating audio SSL models, challenging the reliance on costly fine-tuning.<\/p>","protected":false},"excerpt":{"rendered":"<p>Our paper Unmute the Patch Tokens: Rethinking Probing in Multi-Label Audio Classification by Lukas Rauch, Ren\u00e9 Heinrich, Houtan Ghaffari, Lukas Miklautz, Ilyass Moummad, Bernhard Sick and Christoph Scholz is online on ArXiv. Abstract: Although probing frozen models has become a standard evaluation paradigm, self-supervised learning in audio defaults to fine-tuning. A key reason is that global pooling [\u2026]<\/p>","protected":false},"author":15,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"site-sidebar-layout":"default","site-content-layout":"","ast-site-content-layout":"default","site-content-style":"default","site-sidebar-style":"default","ast-global-header-display":"","ast-banner-title-visibility":"","ast-main-header-display":"","ast-hfb-above-header-display":"","ast-hfb-below-header-display":"","ast-hfb-mobile-header-display":"","site-post-title":"","ast-breadcrumbs-content":"","ast-featured-img":"","footer-sml-layout":"","theme-transparent-header-meta":"","adv-header-id-meta":"","stick-header-meta":"","header-above-stick-meta":"","header-main-stick-meta":"","header-below-stick-meta":"","astra-migrate-meta-layouts":"default","ast-page-background-enabled":"default","ast-page-background-meta":{"desktop":{"background-color":"var(--ast-global-color-4)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"ast-content-background-meta":{"desktop":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"footnotes":""},"categories":[4],"tags":[],"class_list":["post-1635","post","type-post","status-publish","format-standard","hentry","category-aktuelles"],"_links":{"self":[{"href":"https:\/\/deepbirddetect.de\/en\/wp-json\/wp\/v2\/posts\/1635","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/deepbirddetect.de\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/deepbirddetect.de\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/deepbirddetect.de\/en\/wp-json\/wp\/v2\/users\/15"}],"replies":[{"embeddable":true,"href":"https:\/\/deepbirddetect.de\/en\/wp-json\/wp\/v2\/comments?post=1635"}],"version-history":[{"count":1,"href":"https:\/\/deepbirddetect.de\/en\/wp-json\/wp\/v2\/posts\/1635\/revisions"}],"predecessor-version":[{"id":1636,"href":"https:\/\/deepbirddetect.de\/en\/wp-json\/wp\/v2\/posts\/1635\/revisions\/1636"}],"wp:attachment":[{"href":"https:\/\/deepbirddetect.de\/en\/wp-json\/wp\/v2\/media?parent=1635"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/deepbirddetect.de\/en\/wp-json\/wp\/v2\/categories?post=1635"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/deepbirddetect.de\/en\/wp-json\/wp\/v2\/tags?post=1635"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}