{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:N4F2ZOOBPBWIYA3VDAXRC5HT5J","short_pith_number":"pith:N4F2ZOOB","schema_version":"1.0","canonical_sha256":"6f0bacb9c1786c8c0375182f1174f3ea5bb5ae9abe6c5597bf85af8d29669e25","source":{"kind":"arxiv","id":"1910.10288","version":2},"attestation_state":"computed","paper":{"title":"Location-Relative Attention Mechanisms For Robust Long-Form Speech Synthesis","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","cs.SD","eess.AS"],"primary_cat":"cs.CL","authors_text":"Daisy Stanton, David Kao, Eric Battenberg, Matt Shannon, RJ Skerry-Ryan, Soroosh Mariooryad, Tom Bagby","submitted_at":"2019-10-23T00:21:33Z","abstract_excerpt":"Despite the ability to produce human-level speech for in-domain text, attention-based end-to-end text-to-speech (TTS) systems suffer from text alignment failures that increase in frequency for out-of-domain text. We show that these failures can be addressed using simple location-relative attention mechanisms that do away with content-based query/key comparisons. We compare two families of attention mechanisms: location-relative GMM-based mechanisms and additive energy-based mechanisms. We suggest simple modifications to GMM-based attention that allow it to align quickly and consistently during"},"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":"1910.10288","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-10-23T00:21:33Z","cross_cats_sorted":["cs.LG","cs.SD","eess.AS"],"title_canon_sha256":"5ca6ad0ba3225110395b790a214d0ff90ccc3c575023c075ca6e6d15d4002e9d","abstract_canon_sha256":"75fe9dbaf9d83fa789c7550ac26b52793bf6014ee163e53fba4609dfa4c53bd5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:57:34.167352Z","signature_b64":"0HK/w/aUQA6L8FFsuYkcEYRHfS1Gd7tyJcJKUEpIjdpYPuuEp95YPgvFzk2MYKbUiRwMBHpt51Ws+PIrWvfbBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6f0bacb9c1786c8c0375182f1174f3ea5bb5ae9abe6c5597bf85af8d29669e25","last_reissued_at":"2026-07-05T00:57:34.166867Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:57:34.166867Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Location-Relative Attention Mechanisms For Robust Long-Form Speech Synthesis","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","cs.SD","eess.AS"],"primary_cat":"cs.CL","authors_text":"Daisy Stanton, David Kao, Eric Battenberg, Matt Shannon, RJ Skerry-Ryan, Soroosh Mariooryad, Tom Bagby","submitted_at":"2019-10-23T00:21:33Z","abstract_excerpt":"Despite the ability to produce human-level speech for in-domain text, attention-based end-to-end text-to-speech (TTS) systems suffer from text alignment failures that increase in frequency for out-of-domain text. We show that these failures can be addressed using simple location-relative attention mechanisms that do away with content-based query/key comparisons. We compare two families of attention mechanisms: location-relative GMM-based mechanisms and additive energy-based mechanisms. We suggest simple modifications to GMM-based attention that allow it to align quickly and consistently during"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1910.10288","kind":"arxiv","version":2},"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/1910.10288/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":"1910.10288","created_at":"2026-07-05T00:57:34.166926+00:00"},{"alias_kind":"arxiv_version","alias_value":"1910.10288v2","created_at":"2026-07-05T00:57:34.166926+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1910.10288","created_at":"2026-07-05T00:57:34.166926+00:00"},{"alias_kind":"pith_short_12","alias_value":"N4F2ZOOBPBWI","created_at":"2026-07-05T00:57:34.166926+00:00"},{"alias_kind":"pith_short_16","alias_value":"N4F2ZOOBPBWIYA3V","created_at":"2026-07-05T00:57:34.166926+00:00"},{"alias_kind":"pith_short_8","alias_value":"N4F2ZOOB","created_at":"2026-07-05T00:57:34.166926+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2505.16279","citing_title":"MM-MovieDubber: Towards Multi-Modal Learning for Multi-Modal Movie Dubbing","ref_index":31,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/N4F2ZOOBPBWIYA3VDAXRC5HT5J","json":"https://pith.science/pith/N4F2ZOOBPBWIYA3VDAXRC5HT5J.json","graph_json":"https://pith.science/api/pith-number/N4F2ZOOBPBWIYA3VDAXRC5HT5J/graph.json","events_json":"https://pith.science/api/pith-number/N4F2ZOOBPBWIYA3VDAXRC5HT5J/events.json","paper":"https://pith.science/paper/N4F2ZOOB"},"agent_actions":{"view_html":"https://pith.science/pith/N4F2ZOOBPBWIYA3VDAXRC5HT5J","download_json":"https://pith.science/pith/N4F2ZOOBPBWIYA3VDAXRC5HT5J.json","view_paper":"https://pith.science/paper/N4F2ZOOB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1910.10288&json=true","fetch_graph":"https://pith.science/api/pith-number/N4F2ZOOBPBWIYA3VDAXRC5HT5J/graph.json","fetch_events":"https://pith.science/api/pith-number/N4F2ZOOBPBWIYA3VDAXRC5HT5J/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/N4F2ZOOBPBWIYA3VDAXRC5HT5J/action/timestamp_anchor","attest_storage":"https://pith.science/pith/N4F2ZOOBPBWIYA3VDAXRC5HT5J/action/storage_attestation","attest_author":"https://pith.science/pith/N4F2ZOOBPBWIYA3VDAXRC5HT5J/action/author_attestation","sign_citation":"https://pith.science/pith/N4F2ZOOBPBWIYA3VDAXRC5HT5J/action/citation_signature","submit_replication":"https://pith.science/pith/N4F2ZOOBPBWIYA3VDAXRC5HT5J/action/replication_record"}},"created_at":"2026-07-05T00:57:34.166926+00:00","updated_at":"2026-07-05T00:57:34.166926+00:00"}