{"paper":{"title":"Exploring Speech Foundation Models for Speaker Diarization Across Lifespan","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Speech diarization models trained only on adults lose accuracy on child and older-adult conversations but recover with joint multi-age training.","cross_cats":[],"primary_cat":"eess.AS","authors_text":"Anfeng Xu, Shrikanth Narayanan, Tiantian Feng","submitted_at":"2026-04-06T21:57:21Z","abstract_excerpt":"Speech foundation models have shown strong transferability across a wide range of speech applications. However, their robustness to age-related domain shift in speaker diarization remains underexplored. In this work, we present a cross-lifespan evaluation within a unified end-to-end neural diarization framework (EEND-VC), covering speech samples from conversations involving children, adults, and older adults. We compare models under zero-shot cross-age inference, joint multi-age training, and domain-specific adaptation. Results show substantial performance degradation when models trained on ad"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Results show substantial performance degradation when models trained on adult-specific speech are applied to child and older-adult conversational data. Moreover, joint multi-age training across different age groups improves robustness without reducing diarization performance in canonical adult conversations, while targeted age group adaptation yields further gains in diarization performance, particularly when using the Whisper encoder.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That observed performance differences are caused primarily by age-related acoustic domain shift rather than differences in recording quality, conversation style, or speaker count across the chosen datasets.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Adult-trained speech foundation models lose diarization accuracy on child and older-adult speech, but multi-age joint training and adaptation improve robustness across the lifespan.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Speech diarization models trained only on adults lose accuracy on child and older-adult conversations but recover with joint multi-age training.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"4af6e96c49ea4933faecfc1fcc7cf3ed5b72617969bd00ef8ffc3da762d63ded"},"source":{"id":"2604.05201","kind":"arxiv","version":2},"verdict":{"id":"72a6a035-76d5-4ebf-9012-5df3e6e043fb","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-10T18:45:20.643306Z","strongest_claim":"Results show substantial performance degradation when models trained on adult-specific speech are applied to child and older-adult conversational data. Moreover, joint multi-age training across different age groups improves robustness without reducing diarization performance in canonical adult conversations, while targeted age group adaptation yields further gains in diarization performance, particularly when using the Whisper encoder.","one_line_summary":"Adult-trained speech foundation models lose diarization accuracy on child and older-adult speech, but multi-age joint training and adaptation improve robustness across the lifespan.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That observed performance differences are caused primarily by age-related acoustic domain shift rather than differences in recording quality, conversation style, or speaker count across the chosen datasets.","pith_extraction_headline":"Speech diarization models trained only on adults lose accuracy on child and older-adult conversations but recover with joint multi-age training."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.05201/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}