{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:MRVGG5YK547P6NT4563XFDWB3R","short_pith_number":"pith:MRVGG5YK","schema_version":"1.0","canonical_sha256":"646a63770aef3eff367cefb7728ec1dc6bc8c01bb71d50e2783c240995ea2765","source":{"kind":"arxiv","id":"2412.01195","version":1},"attestation_state":"computed","paper":{"title":"Memory-Efficient Training for Deep Speaker Embedding Learning in Speaker Verification","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.SD"],"primary_cat":"eess.AS","authors_text":"Bei Liu, Yanmin Qian","submitted_at":"2024-12-02T06:57:46Z","abstract_excerpt":"Recent speaker verification (SV) systems have shown a trend toward adopting deeper speaker embedding extractors. Although deeper and larger neural networks can significantly improve performance, their substantial memory requirements hinder training on consumer GPUs. In this paper, we explore a memory-efficient training strategy for deep speaker embedding learning in resource-constrained scenarios. Firstly, we conduct a systematic analysis of GPU memory allocation during SV system training. Empirical observations show that activations and optimizer states are the main sources of memory consumpt"},"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":"2412.01195","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"eess.AS","submitted_at":"2024-12-02T06:57:46Z","cross_cats_sorted":["cs.AI","cs.SD"],"title_canon_sha256":"009440bc5baa07cdc639627e3abc4a1f9af3a83061a5a4774eb7754c4eefc86e","abstract_canon_sha256":"6f6f08be0639b4bc03590af1dfdcd5dd7e742123ee2e74e56592aed16ab63ffc"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:42:55.650362Z","signature_b64":"ugAYzGjpjNGeEhCb19MD3tFKLVgoYkqQGLqdDoDD+eITWNRvDkKxrDhvYsht7zxO0E5S4P7gcFuFAETUR2g7Ag==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"646a63770aef3eff367cefb7728ec1dc6bc8c01bb71d50e2783c240995ea2765","last_reissued_at":"2026-07-05T09:42:55.649772Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:42:55.649772Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Memory-Efficient Training for Deep Speaker Embedding Learning in Speaker Verification","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.SD"],"primary_cat":"eess.AS","authors_text":"Bei Liu, Yanmin Qian","submitted_at":"2024-12-02T06:57:46Z","abstract_excerpt":"Recent speaker verification (SV) systems have shown a trend toward adopting deeper speaker embedding extractors. Although deeper and larger neural networks can significantly improve performance, their substantial memory requirements hinder training on consumer GPUs. In this paper, we explore a memory-efficient training strategy for deep speaker embedding learning in resource-constrained scenarios. Firstly, we conduct a systematic analysis of GPU memory allocation during SV system training. Empirical observations show that activations and optimizer states are the main sources of memory consumpt"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.01195","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/2412.01195/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":"2412.01195","created_at":"2026-07-05T09:42:55.649838+00:00"},{"alias_kind":"arxiv_version","alias_value":"2412.01195v1","created_at":"2026-07-05T09:42:55.649838+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.01195","created_at":"2026-07-05T09:42:55.649838+00:00"},{"alias_kind":"pith_short_12","alias_value":"MRVGG5YK547P","created_at":"2026-07-05T09:42:55.649838+00:00"},{"alias_kind":"pith_short_16","alias_value":"MRVGG5YK547P6NT4","created_at":"2026-07-05T09:42:55.649838+00:00"},{"alias_kind":"pith_short_8","alias_value":"MRVGG5YK","created_at":"2026-07-05T09:42:55.649838+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/MRVGG5YK547P6NT4563XFDWB3R","json":"https://pith.science/pith/MRVGG5YK547P6NT4563XFDWB3R.json","graph_json":"https://pith.science/api/pith-number/MRVGG5YK547P6NT4563XFDWB3R/graph.json","events_json":"https://pith.science/api/pith-number/MRVGG5YK547P6NT4563XFDWB3R/events.json","paper":"https://pith.science/paper/MRVGG5YK"},"agent_actions":{"view_html":"https://pith.science/pith/MRVGG5YK547P6NT4563XFDWB3R","download_json":"https://pith.science/pith/MRVGG5YK547P6NT4563XFDWB3R.json","view_paper":"https://pith.science/paper/MRVGG5YK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2412.01195&json=true","fetch_graph":"https://pith.science/api/pith-number/MRVGG5YK547P6NT4563XFDWB3R/graph.json","fetch_events":"https://pith.science/api/pith-number/MRVGG5YK547P6NT4563XFDWB3R/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MRVGG5YK547P6NT4563XFDWB3R/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MRVGG5YK547P6NT4563XFDWB3R/action/storage_attestation","attest_author":"https://pith.science/pith/MRVGG5YK547P6NT4563XFDWB3R/action/author_attestation","sign_citation":"https://pith.science/pith/MRVGG5YK547P6NT4563XFDWB3R/action/citation_signature","submit_replication":"https://pith.science/pith/MRVGG5YK547P6NT4563XFDWB3R/action/replication_record"}},"created_at":"2026-07-05T09:42:55.649838+00:00","updated_at":"2026-07-05T09:42:55.649838+00:00"}