{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:CXLZCIOBUJTEISY5GF6DZTASBN","short_pith_number":"pith:CXLZCIOB","schema_version":"1.0","canonical_sha256":"15d79121c1a266444b1d317c3ccc120b70c57608f8060fefed298700288785ff","source":{"kind":"arxiv","id":"2110.10139","version":2},"attestation_state":"computed","paper":{"title":"Chunked Autoregressive GAN for Conditional Waveform Synthesis","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.SD"],"primary_cat":"eess.AS","authors_text":"Aaron Courville, Kundan Kumar, Max Morrison, Prem Seetharaman, Rithesh Kumar, Yoshua Bengio","submitted_at":"2021-10-19T17:48:12Z","abstract_excerpt":"Conditional waveform synthesis models learn a distribution of audio waveforms given conditioning such as text, mel-spectrograms, or MIDI. These systems employ deep generative models that model the waveform via either sequential (autoregressive) or parallel (non-autoregressive) sampling. Generative adversarial networks (GANs) have become a common choice for non-autoregressive waveform synthesis. However, state-of-the-art GAN-based models produce artifacts when performing mel-spectrogram inversion. In this paper, we demonstrate that these artifacts correspond with an inability for the generator "},"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":"2110.10139","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"eess.AS","submitted_at":"2021-10-19T17:48:12Z","cross_cats_sorted":["cs.SD"],"title_canon_sha256":"ee58f4fc00dc55fe60add9f05c68d8bf9e22f7fc4660a242d41b7c42d8ab2500","abstract_canon_sha256":"8181b7c73568d62a0793d7ae89d6ae44a7ae1d697c805afb9b59a447f6a3fd5a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:01:57.485517Z","signature_b64":"aM8O2o1yh58sgk5hAzMQG6nukvOuOaxb6CUhDYyRcN1Ar8kHHbY8pbwYxIOJEAA1uqvfafZPsfESToyUsDd/CQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"15d79121c1a266444b1d317c3ccc120b70c57608f8060fefed298700288785ff","last_reissued_at":"2026-07-05T04:01:57.485024Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:01:57.485024Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Chunked Autoregressive GAN for Conditional Waveform Synthesis","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.SD"],"primary_cat":"eess.AS","authors_text":"Aaron Courville, Kundan Kumar, Max Morrison, Prem Seetharaman, Rithesh Kumar, Yoshua Bengio","submitted_at":"2021-10-19T17:48:12Z","abstract_excerpt":"Conditional waveform synthesis models learn a distribution of audio waveforms given conditioning such as text, mel-spectrograms, or MIDI. These systems employ deep generative models that model the waveform via either sequential (autoregressive) or parallel (non-autoregressive) sampling. Generative adversarial networks (GANs) have become a common choice for non-autoregressive waveform synthesis. However, state-of-the-art GAN-based models produce artifacts when performing mel-spectrogram inversion. In this paper, we demonstrate that these artifacts correspond with an inability for the generator "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2110.10139","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/2110.10139/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":"2110.10139","created_at":"2026-07-05T04:01:57.485082+00:00"},{"alias_kind":"arxiv_version","alias_value":"2110.10139v2","created_at":"2026-07-05T04:01:57.485082+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2110.10139","created_at":"2026-07-05T04:01:57.485082+00:00"},{"alias_kind":"pith_short_12","alias_value":"CXLZCIOBUJTE","created_at":"2026-07-05T04:01:57.485082+00:00"},{"alias_kind":"pith_short_16","alias_value":"CXLZCIOBUJTEISY5","created_at":"2026-07-05T04:01:57.485082+00:00"},{"alias_kind":"pith_short_8","alias_value":"CXLZCIOB","created_at":"2026-07-05T04:01:57.485082+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.16681","citing_title":"A Survey of Advancing Audio Super-Resolution and Bandwidth Extension from Discriminative to Generative Models","ref_index":45,"is_internal_anchor":false},{"citing_arxiv_id":"2505.24437","citing_title":"SwitchCodec: A High-Fidelity Nerual Audio Codec With Sparse Quantization","ref_index":8,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/CXLZCIOBUJTEISY5GF6DZTASBN","json":"https://pith.science/pith/CXLZCIOBUJTEISY5GF6DZTASBN.json","graph_json":"https://pith.science/api/pith-number/CXLZCIOBUJTEISY5GF6DZTASBN/graph.json","events_json":"https://pith.science/api/pith-number/CXLZCIOBUJTEISY5GF6DZTASBN/events.json","paper":"https://pith.science/paper/CXLZCIOB"},"agent_actions":{"view_html":"https://pith.science/pith/CXLZCIOBUJTEISY5GF6DZTASBN","download_json":"https://pith.science/pith/CXLZCIOBUJTEISY5GF6DZTASBN.json","view_paper":"https://pith.science/paper/CXLZCIOB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2110.10139&json=true","fetch_graph":"https://pith.science/api/pith-number/CXLZCIOBUJTEISY5GF6DZTASBN/graph.json","fetch_events":"https://pith.science/api/pith-number/CXLZCIOBUJTEISY5GF6DZTASBN/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CXLZCIOBUJTEISY5GF6DZTASBN/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CXLZCIOBUJTEISY5GF6DZTASBN/action/storage_attestation","attest_author":"https://pith.science/pith/CXLZCIOBUJTEISY5GF6DZTASBN/action/author_attestation","sign_citation":"https://pith.science/pith/CXLZCIOBUJTEISY5GF6DZTASBN/action/citation_signature","submit_replication":"https://pith.science/pith/CXLZCIOBUJTEISY5GF6DZTASBN/action/replication_record"}},"created_at":"2026-07-05T04:01:57.485082+00:00","updated_at":"2026-07-05T04:01:57.485082+00:00"}