{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:3PK4DJBXN6M2LMIDZGHJPYEHN3","short_pith_number":"pith:3PK4DJBX","schema_version":"1.0","canonical_sha256":"dbd5c1a4376f99a5b103c98e97e0876ee95779634fa0434fd3093c90b302be7c","source":{"kind":"arxiv","id":"2603.04862","version":3},"attestation_state":"computed","paper":{"title":"Focus Then Listen: Exploring Plug-and-Play Audio Enhancer for Noise-Robust Large Audio Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.SD","authors_text":"Han Yin, Jung-Woo Choi, Ting Dang, Yang Xiao, Younghoo Kwon","submitted_at":"2026-03-05T06:35:59Z","abstract_excerpt":"Large audio language models (LALMs) are a class of foundation models for audio understanding. Existing LALMs tend to degrade significantly in real-world noisy acoustic conditions where speech and non-speech sounds interfere. While noise-aware fine-tuning can improve robustness, it requires task-specific noisy data and expensive retraining, limiting scalability. To address this issue, we propose Focus-Then-Listen (FTL), a plug-and-play audio enhancer that improves LALMs' noise robustness. Specifically, FTL first separates the input waveform into speech and non-speech, and a modality router is a"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2603.04862","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SD","submitted_at":"2026-03-05T06:35:59Z","cross_cats_sorted":[],"title_canon_sha256":"1dc0a9e99e73172b3a074664ee134d1b160c23d0b01d50bc6f14f2d0d5125b83","abstract_canon_sha256":"f1e9e692c6efa5651affc7d1b3a0a3075e29e16840ae8335250e7e385d5ba51c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-26T01:03:27.109840Z","signature_b64":"kKMRtwnObFdnY9/u7uh1K87vBo0V0XWnU/70XDgwDCmyVrbZiG/KFXJ4dUE0zqrR1PRQp9ll0DosIaTsVG/hBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dbd5c1a4376f99a5b103c98e97e0876ee95779634fa0434fd3093c90b302be7c","last_reissued_at":"2026-05-26T01:03:27.108885Z","signature_status":"signed_v1","first_computed_at":"2026-05-26T01:03:27.108885Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Focus Then Listen: Exploring Plug-and-Play Audio Enhancer for Noise-Robust Large Audio Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.SD","authors_text":"Han Yin, Jung-Woo Choi, Ting Dang, Yang Xiao, Younghoo Kwon","submitted_at":"2026-03-05T06:35:59Z","abstract_excerpt":"Large audio language models (LALMs) are a class of foundation models for audio understanding. Existing LALMs tend to degrade significantly in real-world noisy acoustic conditions where speech and non-speech sounds interfere. While noise-aware fine-tuning can improve robustness, it requires task-specific noisy data and expensive retraining, limiting scalability. To address this issue, we propose Focus-Then-Listen (FTL), a plug-and-play audio enhancer that improves LALMs' noise robustness. Specifically, FTL first separates the input waveform into speech and non-speech, and a modality router is a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2603.04862","kind":"arxiv","version":3},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2603.04862/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2603.04862","created_at":"2026-05-26T01:03:27.109011+00:00"},{"alias_kind":"arxiv_version","alias_value":"2603.04862v3","created_at":"2026-05-26T01:03:27.109011+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2603.04862","created_at":"2026-05-26T01:03:27.109011+00:00"},{"alias_kind":"pith_short_12","alias_value":"3PK4DJBXN6M2","created_at":"2026-05-26T01:03:27.109011+00:00"},{"alias_kind":"pith_short_16","alias_value":"3PK4DJBXN6M2LMID","created_at":"2026-05-26T01:03:27.109011+00:00"},{"alias_kind":"pith_short_8","alias_value":"3PK4DJBX","created_at":"2026-05-26T01:03:27.109011+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2605.23954","citing_title":"EchoDistill:Alignment Noisy-to-Clean Self-Distillation for Robust Audio LLMs","ref_index":60,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3PK4DJBXN6M2LMIDZGHJPYEHN3","json":"https://pith.science/pith/3PK4DJBXN6M2LMIDZGHJPYEHN3.json","graph_json":"https://pith.science/api/pith-number/3PK4DJBXN6M2LMIDZGHJPYEHN3/graph.json","events_json":"https://pith.science/api/pith-number/3PK4DJBXN6M2LMIDZGHJPYEHN3/events.json","paper":"https://pith.science/paper/3PK4DJBX"},"agent_actions":{"view_html":"https://pith.science/pith/3PK4DJBXN6M2LMIDZGHJPYEHN3","download_json":"https://pith.science/pith/3PK4DJBXN6M2LMIDZGHJPYEHN3.json","view_paper":"https://pith.science/paper/3PK4DJBX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2603.04862&json=true","fetch_graph":"https://pith.science/api/pith-number/3PK4DJBXN6M2LMIDZGHJPYEHN3/graph.json","fetch_events":"https://pith.science/api/pith-number/3PK4DJBXN6M2LMIDZGHJPYEHN3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3PK4DJBXN6M2LMIDZGHJPYEHN3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3PK4DJBXN6M2LMIDZGHJPYEHN3/action/storage_attestation","attest_author":"https://pith.science/pith/3PK4DJBXN6M2LMIDZGHJPYEHN3/action/author_attestation","sign_citation":"https://pith.science/pith/3PK4DJBXN6M2LMIDZGHJPYEHN3/action/citation_signature","submit_replication":"https://pith.science/pith/3PK4DJBXN6M2LMIDZGHJPYEHN3/action/replication_record"}},"created_at":"2026-05-26T01:03:27.109011+00:00","updated_at":"2026-05-26T01:03:27.109011+00:00"}