{"id":1136805,"date":"2025-04-16T11:42:01","date_gmt":"2025-04-16T18:42:01","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/?post_type=msr-research-item&p=1136805"},"modified":"2025-04-17T06:33:40","modified_gmt":"2025-04-17T13:33:40","slug":"distill-mos-a-compact-speech-quality-assessment-model","status":"publish","type":"msr-research-item","link":"https:\/\/www.microsoft.com\/en-us\/research\/publication\/distill-mos-a-compact-speech-quality-assessment-model\/","title":{"rendered":"Distillation and Pruning for Scalable Self-Supervised Representation-Based Speech Quality Assessment"},"content":{"rendered":"

Distill-MOS<\/strong> is a compact and efficient speech quality assessment model learned from a larger speech quality assessment model based on wav2vec2.0 XLS-R embeddings. Model weights and inference code are available on GitHub (opens in new tab)<\/span><\/a>.<\/p>\n

Install Distill-MOS via<\/p>\n

pip install distillmos<\/code><\/p>\n

The work is described in the paper: “Distillation and Pruning for Scalable Self-Supervised Representation-Based Speech Quality Assessment”.<\/p>\n

Abstract:
\nIn this paper, we investigate distillation and pruning methods to reduce model size for non-intrusive speech quality assessment based on self-supervised representations. Our experiments build on XLS-R-SQA, a speech quality assessment model using wav2vec 2.0 XLS-R embeddings. We retrain this model on a large compilation of mean opinion score datasets, encompassing over 100,000 labeled clips. For distillation, using this model as a teacher, we generate pseudo-labels on unlabeled degraded speech signals and train student models of varying sizes. For pruning, we use a data-driven strategy. While data-driven pruning performs better at larger model sizes, distillation on unlabeled data is more effective for smaller model sizes. Distillation can halve the gap between the baseline’s correlation with ground-truth MOS labels and that of the XLS-R-based teacher model, while reducing model size by two orders of magnitude compared to the teacher model.<\/p>\n

\"Table

Table summarizing speech quality estimation results<\/p><\/div>\n","protected":false},"excerpt":{"rendered":"

Distill-MOS is a compact and efficient speech quality assessment model learned from a larger speech quality assessment model based on wav2vec2.0 XLS-R embeddings. Model weights and inference code are available on GitHub. Install Distill-MOS via pip install distillmos The work is described in the paper: “Distillation and Pruning for Scalable Self-Supervised Representation-Based Speech Quality Assessment”. […]<\/p>\n","protected":false},"featured_media":0,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"_classifai_error":"","msr-author-ordering":null,"msr_publishername":"","msr_publisher_other":"","msr_booktitle":"","msr_chapter":"","msr_edition":"","msr_editors":"","msr_how_published":"","msr_isbn":"","msr_issue":"","msr_journal":"","msr_number":"","msr_organization":"","msr_pages_string":"","msr_page_range_start":"","msr_page_range_end":"","msr_series":"","msr_volume":"","msr_copyright":"","msr_conference_name":"","msr_doi":"","msr_arxiv_id":"","msr_s2_paper_id":"","msr_mag_id":"","msr_pubmed_id":"","msr_other_authors":"","msr_other_contributors":"","msr_speaker":"","msr_award":"","msr_affiliation":"","msr_institution":"","msr_host":"","msr_version":"","msr_duration":"","msr_original_fields_of_study":null,"msr_release_tracker_id":"","msr_s2_match_type":"","msr_citation_count_updated":"","msr_published_date":"2025-4-1","msr_highlight_text":"","msr_notes":"","msr_longbiography":"","msr_publicationurl":"","msr_external_url":"","msr_secondary_video_url":"","msr_conference_url":"","msr_journal_url":"","msr_s2_pdf_url":"","msr_year":0,"msr_citation_count":0,"msr_influential_citations":0,"msr_reference_count":0,"msr_s2_match_confidence":0,"msr_microsoftintellectualproperty":true,"msr_s2_open_access":false,"msr_s2_author_ids":[],"msr_pub_ids":[],"msr_hide_image_in_river":null,"footnotes":""},"msr-research-highlight":[],"research-area":[243062],"msr-publication-type":[193716],"msr-publisher":[],"msr-focus-area":[],"msr-locale":[268875],"msr-post-option":[269148,269142],"msr-field-of-study":[264253],"msr-conference":[259657],"msr-journal":[],"msr-impact-theme":[],"msr-pillar":[],"class_list":["post-1136805","msr-research-item","type-msr-research-item","status-publish","hentry","msr-research-area-audio-acoustics","msr-locale-en_us","msr-post-option-approved-for-river","msr-post-option-include-in-river","msr-field-of-study-audio-and-speech-processing"],"msr_publishername":"","msr_edition":"","msr_affiliation":"","msr_published_date":"2025-4-1","msr_host":"","msr_duration":"","msr_version":"","msr_speaker":"","msr_other_contributors":"","msr_booktitle":"","msr_pages_string":"","msr_chapter":"","msr_isbn":"","msr_journal":"","msr_volume":"","msr_number":"","msr_editors":"","msr_series":"","msr_issue":"","msr_organization":"","msr_how_published":"","msr_notes":"","msr_highlight_text":"","msr_release_tracker_id":"","msr_original_fields_of_study":"","msr_download_urls":"","msr_external_url":"","msr_secondary_video_url":"","msr_longbiography":"","msr_microsoftintellectualproperty":1,"msr_main_download":"","msr_publicationurl":"","msr_doi":"","msr_publication_uploader":[{"type":"url","viewUrl":"false","id":"false","title":"https:\/\/arxiv.org\/pdf\/2502.05356","label_id":"243132","label":0}],"msr_related_uploader":[{"type":"url","viewUrl":"false","id":"false","title":"https:\/\/github.com\/microsoft\/Distill-MOS","label_id":"264520","label":0}],"msr_citation_count":0,"msr_citation_count_updated":"","msr_s2_paper_id":"","msr_influential_citations":0,"msr_reference_count":0,"msr_arxiv_id":"","msr_s2_author_ids":[],"msr_s2_open_access":false,"msr_s2_pdf_url":null,"msr_attachments":[],"msr-author-ordering":[{"type":"text","value":"Benjamin Stahl","user_id":0,"rest_url":false},{"type":"user_nicename","value":"Hannes Gamper","user_id":31943,"rest_url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=Hannes Gamper"}],"msr_impact_theme":[],"msr_research_lab":[199565],"msr_event":[],"msr_group":[144923],"msr_project":[],"publication":[],"video":[],"msr-tool":[],"msr_publication_type":"inproceedings","related_content":[],"_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/1136805","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item"}],"about":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-research-item"}],"version-history":[{"count":8,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/1136805\/revisions"}],"predecessor-version":[{"id":1136868,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/1136805\/revisions\/1136868"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=1136805"}],"wp:term":[{"taxonomy":"msr-research-highlight","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-highlight?post=1136805"},{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=1136805"},{"taxonomy":"msr-publication-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-publication-type?post=1136805"},{"taxonomy":"msr-publisher","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-publisher?post=1136805"},{"taxonomy":"msr-focus-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-focus-area?post=1136805"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=1136805"},{"taxonomy":"msr-post-option","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-post-option?post=1136805"},{"taxonomy":"msr-field-of-study","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-field-of-study?post=1136805"},{"taxonomy":"msr-conference","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-conference?post=1136805"},{"taxonomy":"msr-journal","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-journal?post=1136805"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=1136805"},{"taxonomy":"msr-pillar","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-pillar?post=1136805"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}