{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:DF55HVOZWC5CVPPPXXSCG2TLZ7","short_pith_number":"pith:DF55HVOZ","schema_version":"1.0","canonical_sha256":"197bd3d5d9b0ba2abdefbde4236a6bcffe5719dd006550323c66177a29d19fee","source":{"kind":"arxiv","id":"2111.03600","version":3},"attestation_state":"computed","paper":{"title":"Hybrid Spectrogram and Waveform Source Separation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.SD","stat.ML"],"primary_cat":"eess.AS","authors_text":"Alexandre D\\'efossez","submitted_at":"2021-11-05T16:37:45Z","abstract_excerpt":"Source separation models either work on the spectrogram or waveform domain. In this work, we show how to perform end-to-end hybrid source separation, letting the model decide which domain is best suited for each source, and even combining both. The proposed hybrid version of the Demucs architecture won the Music Demixing Challenge 2021 organized by Sony. This architecture also comes with additional improvements, such as compressed residual branches, local attention or singular value regularization. Overall, a 1.4 dB improvement of the Signal-To-Distortion (SDR) was observed across all sources "},"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":"2111.03600","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"eess.AS","submitted_at":"2021-11-05T16:37:45Z","cross_cats_sorted":["cs.SD","stat.ML"],"title_canon_sha256":"8b0112ac9e55ce0ce411ef9354e313c2ece0a50d6ea203ac1bd11e115b516c14","abstract_canon_sha256":"f16b93eb353f227a11784ee2618cce67e02c0115d34966a399994c39242fa9de"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:52:36.855603Z","signature_b64":"jd++AnSbISfgvd9RmsGectMI3qOtaY9/c6tWg0T8jF6CmaXj09XAJReu4h88XGV6G6o1nHTODCeBMtQBUS4dCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"197bd3d5d9b0ba2abdefbde4236a6bcffe5719dd006550323c66177a29d19fee","last_reissued_at":"2026-07-05T04:52:36.855191Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:52:36.855191Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Hybrid Spectrogram and Waveform Source Separation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.SD","stat.ML"],"primary_cat":"eess.AS","authors_text":"Alexandre D\\'efossez","submitted_at":"2021-11-05T16:37:45Z","abstract_excerpt":"Source separation models either work on the spectrogram or waveform domain. In this work, we show how to perform end-to-end hybrid source separation, letting the model decide which domain is best suited for each source, and even combining both. The proposed hybrid version of the Demucs architecture won the Music Demixing Challenge 2021 organized by Sony. This architecture also comes with additional improvements, such as compressed residual branches, local attention or singular value regularization. Overall, a 1.4 dB improvement of the Signal-To-Distortion (SDR) was observed across all sources "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2111.03600","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/2111.03600/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":"2111.03600","created_at":"2026-07-05T04:52:36.855248+00:00"},{"alias_kind":"arxiv_version","alias_value":"2111.03600v3","created_at":"2026-07-05T04:52:36.855248+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2111.03600","created_at":"2026-07-05T04:52:36.855248+00:00"},{"alias_kind":"pith_short_12","alias_value":"DF55HVOZWC5C","created_at":"2026-07-05T04:52:36.855248+00:00"},{"alias_kind":"pith_short_16","alias_value":"DF55HVOZWC5CVPPP","created_at":"2026-07-05T04:52:36.855248+00:00"},{"alias_kind":"pith_short_8","alias_value":"DF55HVOZ","created_at":"2026-07-05T04:52:36.855248+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.21661","citing_title":"UnityShots: Memory-Driven Multi-Shot Audio-Video Generation with Boundary-Aware Gating","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2606.01686","citing_title":"HAIM: Human-AI Music Datasets for AI Music Production Tracking Benchmark","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2604.09371","citing_title":"Discrete Token Modeling for Multi-Stem Music Source Separation with Language Models","ref_index":10,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DF55HVOZWC5CVPPPXXSCG2TLZ7","json":"https://pith.science/pith/DF55HVOZWC5CVPPPXXSCG2TLZ7.json","graph_json":"https://pith.science/api/pith-number/DF55HVOZWC5CVPPPXXSCG2TLZ7/graph.json","events_json":"https://pith.science/api/pith-number/DF55HVOZWC5CVPPPXXSCG2TLZ7/events.json","paper":"https://pith.science/paper/DF55HVOZ"},"agent_actions":{"view_html":"https://pith.science/pith/DF55HVOZWC5CVPPPXXSCG2TLZ7","download_json":"https://pith.science/pith/DF55HVOZWC5CVPPPXXSCG2TLZ7.json","view_paper":"https://pith.science/paper/DF55HVOZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2111.03600&json=true","fetch_graph":"https://pith.science/api/pith-number/DF55HVOZWC5CVPPPXXSCG2TLZ7/graph.json","fetch_events":"https://pith.science/api/pith-number/DF55HVOZWC5CVPPPXXSCG2TLZ7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DF55HVOZWC5CVPPPXXSCG2TLZ7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DF55HVOZWC5CVPPPXXSCG2TLZ7/action/storage_attestation","attest_author":"https://pith.science/pith/DF55HVOZWC5CVPPPXXSCG2TLZ7/action/author_attestation","sign_citation":"https://pith.science/pith/DF55HVOZWC5CVPPPXXSCG2TLZ7/action/citation_signature","submit_replication":"https://pith.science/pith/DF55HVOZWC5CVPPPXXSCG2TLZ7/action/replication_record"}},"created_at":"2026-07-05T04:52:36.855248+00:00","updated_at":"2026-07-05T04:52:36.855248+00:00"}