{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:QZZZN44SQZGDR4F7ZB5IVHPUOR","short_pith_number":"pith:QZZZN44S","schema_version":"1.0","canonical_sha256":"867396f392864c38f0bfc87a8a9df4745c4a8c34c1102b5bbee95af46a2163d7","source":{"kind":"arxiv","id":"2502.21291","version":4},"attestation_state":"computed","paper":{"title":"MIGE: Mutually Enhanced Multimodal Instruction-Based Image Generation and Editing","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bingbing Xu, Huawei Shen, Wei Li, Xueyun Tian, Yige Yuan, Yuanzhuo Wang","submitted_at":"2025-02-28T18:21:08Z","abstract_excerpt":"Despite significant progress in diffusion-based image generation, subject-driven generation and instruction-based editing remain challenging. Existing methods typically treat them separately, struggling with limited high-quality data and poor generalization. However, both tasks require capturing complex visual variations while maintaining consistency between inputs and outputs. Inspired by this, we propose MIGE, a unified framework that standardizes task representations using multimodal instructions. It first treats subject-driven generation as creation on a blank canvas and instruction-based "},"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":"2502.21291","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-02-28T18:21:08Z","cross_cats_sorted":[],"title_canon_sha256":"0347f15d39a4c26d01b08c939c4f7ee621e6c635ca7a47cbf453f487951bf15e","abstract_canon_sha256":"739a27aed224136ec542ea1a2a4307442275e1a89ad0fbbecd1426f18320624f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:44:01.742441Z","signature_b64":"lNUb2RjTQUX8gE8iGfW1TP3sBqNaqO0+2QU+SM00aELq/nI8I5FQd0zlxbVXjkZ1b/OsF+2jH2R5nEWIFkykCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"867396f392864c38f0bfc87a8a9df4745c4a8c34c1102b5bbee95af46a2163d7","last_reissued_at":"2026-07-05T11:44:01.741975Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:44:01.741975Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MIGE: Mutually Enhanced Multimodal Instruction-Based Image Generation and Editing","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bingbing Xu, Huawei Shen, Wei Li, Xueyun Tian, Yige Yuan, Yuanzhuo Wang","submitted_at":"2025-02-28T18:21:08Z","abstract_excerpt":"Despite significant progress in diffusion-based image generation, subject-driven generation and instruction-based editing remain challenging. Existing methods typically treat them separately, struggling with limited high-quality data and poor generalization. However, both tasks require capturing complex visual variations while maintaining consistency between inputs and outputs. Inspired by this, we propose MIGE, a unified framework that standardizes task representations using multimodal instructions. It first treats subject-driven generation as creation on a blank canvas and instruction-based "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.21291","kind":"arxiv","version":4},"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/2502.21291/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":"2502.21291","created_at":"2026-07-05T11:44:01.742038+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.21291v4","created_at":"2026-07-05T11:44:01.742038+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.21291","created_at":"2026-07-05T11:44:01.742038+00:00"},{"alias_kind":"pith_short_12","alias_value":"QZZZN44SQZGD","created_at":"2026-07-05T11:44:01.742038+00:00"},{"alias_kind":"pith_short_16","alias_value":"QZZZN44SQZGDR4F7","created_at":"2026-07-05T11:44:01.742038+00:00"},{"alias_kind":"pith_short_8","alias_value":"QZZZN44S","created_at":"2026-07-05T11:44:01.742038+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2506.18871","citing_title":"OmniGen2: Towards Instruction-Aligned Multimodal Generation","ref_index":73,"is_internal_anchor":false},{"citing_arxiv_id":"2507.01908","citing_title":"Reasoning to Edit: Hypothetical Instruction-Based Image Editing with Visual Reasoning","ref_index":17,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QZZZN44SQZGDR4F7ZB5IVHPUOR","json":"https://pith.science/pith/QZZZN44SQZGDR4F7ZB5IVHPUOR.json","graph_json":"https://pith.science/api/pith-number/QZZZN44SQZGDR4F7ZB5IVHPUOR/graph.json","events_json":"https://pith.science/api/pith-number/QZZZN44SQZGDR4F7ZB5IVHPUOR/events.json","paper":"https://pith.science/paper/QZZZN44S"},"agent_actions":{"view_html":"https://pith.science/pith/QZZZN44SQZGDR4F7ZB5IVHPUOR","download_json":"https://pith.science/pith/QZZZN44SQZGDR4F7ZB5IVHPUOR.json","view_paper":"https://pith.science/paper/QZZZN44S","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.21291&json=true","fetch_graph":"https://pith.science/api/pith-number/QZZZN44SQZGDR4F7ZB5IVHPUOR/graph.json","fetch_events":"https://pith.science/api/pith-number/QZZZN44SQZGDR4F7ZB5IVHPUOR/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QZZZN44SQZGDR4F7ZB5IVHPUOR/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QZZZN44SQZGDR4F7ZB5IVHPUOR/action/storage_attestation","attest_author":"https://pith.science/pith/QZZZN44SQZGDR4F7ZB5IVHPUOR/action/author_attestation","sign_citation":"https://pith.science/pith/QZZZN44SQZGDR4F7ZB5IVHPUOR/action/citation_signature","submit_replication":"https://pith.science/pith/QZZZN44SQZGDR4F7ZB5IVHPUOR/action/replication_record"}},"created_at":"2026-07-05T11:44:01.742038+00:00","updated_at":"2026-07-05T11:44:01.742038+00:00"}