{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:KFYPPIWRMPHQRXPLRLOTOVYN3B","short_pith_number":"pith:KFYPPIWR","schema_version":"1.0","canonical_sha256":"5170f7a2d163cf08ddeb8add37570dd8705640a87174f058f695daf5f110a2b9","source":{"kind":"arxiv","id":"2110.05909","version":3},"attestation_state":"computed","paper":{"title":"Rescoring Sequence-to-Sequence Models for Text Line Recognition with CTC-Prefixes","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Christoph Wick, Jochen Z\\\"ollner, Tobias Gr\\\"uning","submitted_at":"2021-10-12T11:40:05Z","abstract_excerpt":"In contrast to Connectionist Temporal Classification (CTC) approaches, Sequence-To-Sequence (S2S) models for Handwritten Text Recognition (HTR) suffer from errors such as skipped or repeated words which often occur at the end of a sequence. In this paper, to combine the best of both approaches, we propose to use the CTC-Prefix-Score during S2S decoding. Hereby, during beam search, paths that are invalid according to the CTC confidence matrix are penalised. Our network architecture is composed of a Convolutional Neural Network (CNN) as visual backbone, bidirectional Long-Short-Term-Memory-Cells"},"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":"2110.05909","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2021-10-12T11:40:05Z","cross_cats_sorted":[],"title_canon_sha256":"1e43549782e3e87bde8d555e2b54e3857353b5cac3e9c86e415b3f648eb8abe7","abstract_canon_sha256":"3e34e0241f0a367ccb2fa42587c16e076d9d9f6d81f9f78c56be470b3e5e3ff0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:09:15.760844Z","signature_b64":"K54aVFyFQ7gETpJFbCeQoBmH2JNbdMBjBxptE7+fWG423smAeU4vTKX423htPd21yl4U9z5TwT2RD+bS/po4AA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5170f7a2d163cf08ddeb8add37570dd8705640a87174f058f695daf5f110a2b9","last_reissued_at":"2026-07-05T04:09:15.760395Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:09:15.760395Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Rescoring Sequence-to-Sequence Models for Text Line Recognition with CTC-Prefixes","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Christoph Wick, Jochen Z\\\"ollner, Tobias Gr\\\"uning","submitted_at":"2021-10-12T11:40:05Z","abstract_excerpt":"In contrast to Connectionist Temporal Classification (CTC) approaches, Sequence-To-Sequence (S2S) models for Handwritten Text Recognition (HTR) suffer from errors such as skipped or repeated words which often occur at the end of a sequence. In this paper, to combine the best of both approaches, we propose to use the CTC-Prefix-Score during S2S decoding. Hereby, during beam search, paths that are invalid according to the CTC confidence matrix are penalised. Our network architecture is composed of a Convolutional Neural Network (CNN) as visual backbone, bidirectional Long-Short-Term-Memory-Cells"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2110.05909","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/2110.05909/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":"2110.05909","created_at":"2026-07-05T04:09:15.760456+00:00"},{"alias_kind":"arxiv_version","alias_value":"2110.05909v3","created_at":"2026-07-05T04:09:15.760456+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2110.05909","created_at":"2026-07-05T04:09:15.760456+00:00"},{"alias_kind":"pith_short_12","alias_value":"KFYPPIWRMPHQ","created_at":"2026-07-05T04:09:15.760456+00:00"},{"alias_kind":"pith_short_16","alias_value":"KFYPPIWRMPHQRXPL","created_at":"2026-07-05T04:09:15.760456+00:00"},{"alias_kind":"pith_short_8","alias_value":"KFYPPIWR","created_at":"2026-07-05T04:09:15.760456+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2508.09936","citing_title":"Quo Vadis Handwritten Text Generation for Handwritten Text Recognition?","ref_index":67,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/KFYPPIWRMPHQRXPLRLOTOVYN3B","json":"https://pith.science/pith/KFYPPIWRMPHQRXPLRLOTOVYN3B.json","graph_json":"https://pith.science/api/pith-number/KFYPPIWRMPHQRXPLRLOTOVYN3B/graph.json","events_json":"https://pith.science/api/pith-number/KFYPPIWRMPHQRXPLRLOTOVYN3B/events.json","paper":"https://pith.science/paper/KFYPPIWR"},"agent_actions":{"view_html":"https://pith.science/pith/KFYPPIWRMPHQRXPLRLOTOVYN3B","download_json":"https://pith.science/pith/KFYPPIWRMPHQRXPLRLOTOVYN3B.json","view_paper":"https://pith.science/paper/KFYPPIWR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2110.05909&json=true","fetch_graph":"https://pith.science/api/pith-number/KFYPPIWRMPHQRXPLRLOTOVYN3B/graph.json","fetch_events":"https://pith.science/api/pith-number/KFYPPIWRMPHQRXPLRLOTOVYN3B/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/KFYPPIWRMPHQRXPLRLOTOVYN3B/action/timestamp_anchor","attest_storage":"https://pith.science/pith/KFYPPIWRMPHQRXPLRLOTOVYN3B/action/storage_attestation","attest_author":"https://pith.science/pith/KFYPPIWRMPHQRXPLRLOTOVYN3B/action/author_attestation","sign_citation":"https://pith.science/pith/KFYPPIWRMPHQRXPLRLOTOVYN3B/action/citation_signature","submit_replication":"https://pith.science/pith/KFYPPIWRMPHQRXPLRLOTOVYN3B/action/replication_record"}},"created_at":"2026-07-05T04:09:15.760456+00:00","updated_at":"2026-07-05T04:09:15.760456+00:00"}