{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:S54WALETV766ZKTKO6TRYWLEO7","short_pith_number":"pith:S54WALET","schema_version":"1.0","canonical_sha256":"9779602c93affdecaa6a77a71c596477f292256b257171e83d6374d035b23b79","source":{"kind":"arxiv","id":"2308.04553","version":3},"attestation_state":"computed","paper":{"title":"From Fake to Real: Pretraining on Balanced Synthetic Images to Prevent Spurious Correlations in Image Recognition","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Bryan A. Plummer, Kate Saenko, Maan Qraitem","submitted_at":"2023-08-08T19:52:28Z","abstract_excerpt":"Visual recognition models are prone to learning spurious correlations induced by a biased training set where certain conditions $B$ (\\eg, Indoors) are over-represented in certain classes $Y$ (\\eg, Big Dogs). Synthetic data from off-the-shelf large-scale generative models offers a promising direction to mitigate this issue by augmenting underrepresented subgroups in the real dataset. However, by using a mixed distribution of real and synthetic data, we introduce another source of bias due to distributional differences between synthetic and real data (\\eg synthetic artifacts). As we will show, p"},"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":"2308.04553","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2023-08-08T19:52:28Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"691dfecb8d2d6a8a5e72da4a78b5743958401e95e3a54247e7e9e9534f705c12","abstract_canon_sha256":"89604a32755655974a89fc4c3b8f2ffae5f5b4a0710a593fa3608ebae570a6a7"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:44:47.043447Z","signature_b64":"5/NDQc3TelIbPiosg2CdOGhO1pJnR8il58lLBMS9oizC6kT3wL9OrXvASeo4vmb4F1npdp1pLZw3LF1Z+VJwDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9779602c93affdecaa6a77a71c596477f292256b257171e83d6374d035b23b79","last_reissued_at":"2026-07-05T08:44:47.042932Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:44:47.042932Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"From Fake to Real: Pretraining on Balanced Synthetic Images to Prevent Spurious Correlations in Image Recognition","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Bryan A. Plummer, Kate Saenko, Maan Qraitem","submitted_at":"2023-08-08T19:52:28Z","abstract_excerpt":"Visual recognition models are prone to learning spurious correlations induced by a biased training set where certain conditions $B$ (\\eg, Indoors) are over-represented in certain classes $Y$ (\\eg, Big Dogs). Synthetic data from off-the-shelf large-scale generative models offers a promising direction to mitigate this issue by augmenting underrepresented subgroups in the real dataset. However, by using a mixed distribution of real and synthetic data, we introduce another source of bias due to distributional differences between synthetic and real data (\\eg synthetic artifacts). As we will show, p"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2308.04553","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/2308.04553/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":"2308.04553","created_at":"2026-07-05T08:44:47.042992+00:00"},{"alias_kind":"arxiv_version","alias_value":"2308.04553v3","created_at":"2026-07-05T08:44:47.042992+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2308.04553","created_at":"2026-07-05T08:44:47.042992+00:00"},{"alias_kind":"pith_short_12","alias_value":"S54WALETV766","created_at":"2026-07-05T08:44:47.042992+00:00"},{"alias_kind":"pith_short_16","alias_value":"S54WALETV766ZKTK","created_at":"2026-07-05T08:44:47.042992+00:00"},{"alias_kind":"pith_short_8","alias_value":"S54WALET","created_at":"2026-07-05T08:44:47.042992+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/S54WALETV766ZKTKO6TRYWLEO7","json":"https://pith.science/pith/S54WALETV766ZKTKO6TRYWLEO7.json","graph_json":"https://pith.science/api/pith-number/S54WALETV766ZKTKO6TRYWLEO7/graph.json","events_json":"https://pith.science/api/pith-number/S54WALETV766ZKTKO6TRYWLEO7/events.json","paper":"https://pith.science/paper/S54WALET"},"agent_actions":{"view_html":"https://pith.science/pith/S54WALETV766ZKTKO6TRYWLEO7","download_json":"https://pith.science/pith/S54WALETV766ZKTKO6TRYWLEO7.json","view_paper":"https://pith.science/paper/S54WALET","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2308.04553&json=true","fetch_graph":"https://pith.science/api/pith-number/S54WALETV766ZKTKO6TRYWLEO7/graph.json","fetch_events":"https://pith.science/api/pith-number/S54WALETV766ZKTKO6TRYWLEO7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/S54WALETV766ZKTKO6TRYWLEO7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/S54WALETV766ZKTKO6TRYWLEO7/action/storage_attestation","attest_author":"https://pith.science/pith/S54WALETV766ZKTKO6TRYWLEO7/action/author_attestation","sign_citation":"https://pith.science/pith/S54WALETV766ZKTKO6TRYWLEO7/action/citation_signature","submit_replication":"https://pith.science/pith/S54WALETV766ZKTKO6TRYWLEO7/action/replication_record"}},"created_at":"2026-07-05T08:44:47.042992+00:00","updated_at":"2026-07-05T08:44:47.042992+00:00"}