{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:H2YXMKLLEURNS6LWTVVGKJBUCG","short_pith_number":"pith:H2YXMKLL","schema_version":"1.0","canonical_sha256":"3eb176296b2522d979769d6a65243411b292d6f4bf58c5750eaa3b63f5c6a183","source":{"kind":"arxiv","id":"2312.10952","version":1},"attestation_state":"computed","paper":{"title":"Soft Alignment of Modality Space for End-to-end Speech Translation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.SD","eess.AS"],"primary_cat":"cs.CL","authors_text":"Bei Li, Chen Xu, Chunliang Zhang, Jingbo Zhu, Kaiqi Kou, Tong Xiao, Yuhao Zhang","submitted_at":"2023-12-18T06:08:51Z","abstract_excerpt":"End-to-end Speech Translation (ST) aims to convert speech into target text within a unified model. The inherent differences between speech and text modalities often impede effective cross-modal and cross-lingual transfer. Existing methods typically employ hard alignment (H-Align) of individual speech and text segments, which can degrade textual representations. To address this, we introduce Soft Alignment (S-Align), using adversarial training to align the representation spaces of both modalities. S-Align creates a modality-invariant space while preserving individual modality quality. Experimen"},"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":"2312.10952","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-12-18T06:08:51Z","cross_cats_sorted":["cs.AI","cs.SD","eess.AS"],"title_canon_sha256":"1c32385a3164ea992dda2cd29e689908660696e5356c99404f8889e5fd8f706c","abstract_canon_sha256":"19743567d63987fa29558f6b53d325a80b2f091e609160decfe91da7e2320443"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:25:14.907177Z","signature_b64":"vm4n2tDYD6M/m/StPRpa7ddAqvDo4LXaMMBcd96rfagAwExYYvEwA0hRgTMXnPE/qzfVnnBRjaU//8VJ6jFoCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3eb176296b2522d979769d6a65243411b292d6f4bf58c5750eaa3b63f5c6a183","last_reissued_at":"2026-07-05T07:25:14.906762Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:25:14.906762Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Soft Alignment of Modality Space for End-to-end Speech Translation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.SD","eess.AS"],"primary_cat":"cs.CL","authors_text":"Bei Li, Chen Xu, Chunliang Zhang, Jingbo Zhu, Kaiqi Kou, Tong Xiao, Yuhao Zhang","submitted_at":"2023-12-18T06:08:51Z","abstract_excerpt":"End-to-end Speech Translation (ST) aims to convert speech into target text within a unified model. The inherent differences between speech and text modalities often impede effective cross-modal and cross-lingual transfer. Existing methods typically employ hard alignment (H-Align) of individual speech and text segments, which can degrade textual representations. To address this, we introduce Soft Alignment (S-Align), using adversarial training to align the representation spaces of both modalities. S-Align creates a modality-invariant space while preserving individual modality quality. Experimen"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2312.10952","kind":"arxiv","version":1},"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/2312.10952/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":"2312.10952","created_at":"2026-07-05T07:25:14.906824+00:00"},{"alias_kind":"arxiv_version","alias_value":"2312.10952v1","created_at":"2026-07-05T07:25:14.906824+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2312.10952","created_at":"2026-07-05T07:25:14.906824+00:00"},{"alias_kind":"pith_short_12","alias_value":"H2YXMKLLEURN","created_at":"2026-07-05T07:25:14.906824+00:00"},{"alias_kind":"pith_short_16","alias_value":"H2YXMKLLEURNS6LW","created_at":"2026-07-05T07:25:14.906824+00:00"},{"alias_kind":"pith_short_8","alias_value":"H2YXMKLL","created_at":"2026-07-05T07:25:14.906824+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/H2YXMKLLEURNS6LWTVVGKJBUCG","json":"https://pith.science/pith/H2YXMKLLEURNS6LWTVVGKJBUCG.json","graph_json":"https://pith.science/api/pith-number/H2YXMKLLEURNS6LWTVVGKJBUCG/graph.json","events_json":"https://pith.science/api/pith-number/H2YXMKLLEURNS6LWTVVGKJBUCG/events.json","paper":"https://pith.science/paper/H2YXMKLL"},"agent_actions":{"view_html":"https://pith.science/pith/H2YXMKLLEURNS6LWTVVGKJBUCG","download_json":"https://pith.science/pith/H2YXMKLLEURNS6LWTVVGKJBUCG.json","view_paper":"https://pith.science/paper/H2YXMKLL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2312.10952&json=true","fetch_graph":"https://pith.science/api/pith-number/H2YXMKLLEURNS6LWTVVGKJBUCG/graph.json","fetch_events":"https://pith.science/api/pith-number/H2YXMKLLEURNS6LWTVVGKJBUCG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/H2YXMKLLEURNS6LWTVVGKJBUCG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/H2YXMKLLEURNS6LWTVVGKJBUCG/action/storage_attestation","attest_author":"https://pith.science/pith/H2YXMKLLEURNS6LWTVVGKJBUCG/action/author_attestation","sign_citation":"https://pith.science/pith/H2YXMKLLEURNS6LWTVVGKJBUCG/action/citation_signature","submit_replication":"https://pith.science/pith/H2YXMKLLEURNS6LWTVVGKJBUCG/action/replication_record"}},"created_at":"2026-07-05T07:25:14.906824+00:00","updated_at":"2026-07-05T07:25:14.906824+00:00"}