{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:3VNTYY2A25BLYA473IVH7C3M26","short_pith_number":"pith:3VNTYY2A","schema_version":"1.0","canonical_sha256":"dd5b3c6340d742bc039fda2a7f8b6cd7bd04255435bae4322328102eeead2218","source":{"kind":"arxiv","id":"2412.07767","version":3},"attestation_state":"computed","paper":{"title":"Learning Visual Generative Priors without Text","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Biao Gong, Chen-Wei Xie, Fan Lu, Jiapeng Zhu, Kecheng Zheng, Shuailei Ma, Wei Wu, Yifei Zhang, Ying Wei, Yujun Shen","submitted_at":"2024-12-10T18:59:31Z","abstract_excerpt":"Although text-to-image (T2I) models have recently thrived as visual generative priors, their reliance on high-quality text-image pairs makes scaling up expensive. We argue that grasping the cross-modality alignment is not a necessity for a sound visual generative prior, whose focus should be on texture modeling. Such a philosophy inspires us to study image-to-image (I2I) generation, where models can learn from in-the-wild images in a self-supervised manner. We first develop a pure vision-based training framework, Lumos, and confirm the feasibility and the scalability of learning I2I models. We"},"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":"2412.07767","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-12-10T18:59:31Z","cross_cats_sorted":[],"title_canon_sha256":"15930e87fa97519048c9a989768f19a5ffb559c8e61afdc47f6bc5e4ece706f9","abstract_canon_sha256":"21e83f5e5b696770cb271153cb715dd8017fae0ce35971a36a14af36fd76126b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:37:39.893167Z","signature_b64":"y78+RQ4hG9t92yqk/Nd8L0659duaci4R1UYUlUMFTKV+5irKzHMFu+iAwgKzhOELPda34ibhKDbf5CmscD5KBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dd5b3c6340d742bc039fda2a7f8b6cd7bd04255435bae4322328102eeead2218","last_reissued_at":"2026-07-05T10:37:39.892433Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:37:39.892433Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Learning Visual Generative Priors without Text","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Biao Gong, Chen-Wei Xie, Fan Lu, Jiapeng Zhu, Kecheng Zheng, Shuailei Ma, Wei Wu, Yifei Zhang, Ying Wei, Yujun Shen","submitted_at":"2024-12-10T18:59:31Z","abstract_excerpt":"Although text-to-image (T2I) models have recently thrived as visual generative priors, their reliance on high-quality text-image pairs makes scaling up expensive. We argue that grasping the cross-modality alignment is not a necessity for a sound visual generative prior, whose focus should be on texture modeling. Such a philosophy inspires us to study image-to-image (I2I) generation, where models can learn from in-the-wild images in a self-supervised manner. We first develop a pure vision-based training framework, Lumos, and confirm the feasibility and the scalability of learning I2I models. We"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.07767","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/2412.07767/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":"2412.07767","created_at":"2026-07-05T10:37:39.892520+00:00"},{"alias_kind":"arxiv_version","alias_value":"2412.07767v3","created_at":"2026-07-05T10:37:39.892520+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.07767","created_at":"2026-07-05T10:37:39.892520+00:00"},{"alias_kind":"pith_short_12","alias_value":"3VNTYY2A25BL","created_at":"2026-07-05T10:37:39.892520+00:00"},{"alias_kind":"pith_short_16","alias_value":"3VNTYY2A25BLYA47","created_at":"2026-07-05T10:37:39.892520+00:00"},{"alias_kind":"pith_short_8","alias_value":"3VNTYY2A","created_at":"2026-07-05T10:37:39.892520+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2505.23493","citing_title":"R2I-Bench: Benchmarking Reasoning-Driven Text-to-Image Generation","ref_index":29,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3VNTYY2A25BLYA473IVH7C3M26","json":"https://pith.science/pith/3VNTYY2A25BLYA473IVH7C3M26.json","graph_json":"https://pith.science/api/pith-number/3VNTYY2A25BLYA473IVH7C3M26/graph.json","events_json":"https://pith.science/api/pith-number/3VNTYY2A25BLYA473IVH7C3M26/events.json","paper":"https://pith.science/paper/3VNTYY2A"},"agent_actions":{"view_html":"https://pith.science/pith/3VNTYY2A25BLYA473IVH7C3M26","download_json":"https://pith.science/pith/3VNTYY2A25BLYA473IVH7C3M26.json","view_paper":"https://pith.science/paper/3VNTYY2A","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2412.07767&json=true","fetch_graph":"https://pith.science/api/pith-number/3VNTYY2A25BLYA473IVH7C3M26/graph.json","fetch_events":"https://pith.science/api/pith-number/3VNTYY2A25BLYA473IVH7C3M26/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3VNTYY2A25BLYA473IVH7C3M26/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3VNTYY2A25BLYA473IVH7C3M26/action/storage_attestation","attest_author":"https://pith.science/pith/3VNTYY2A25BLYA473IVH7C3M26/action/author_attestation","sign_citation":"https://pith.science/pith/3VNTYY2A25BLYA473IVH7C3M26/action/citation_signature","submit_replication":"https://pith.science/pith/3VNTYY2A25BLYA473IVH7C3M26/action/replication_record"}},"created_at":"2026-07-05T10:37:39.892520+00:00","updated_at":"2026-07-05T10:37:39.892520+00:00"}