{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:KDB5WWRXHHKKMC6NATE3LJ5V3Z","short_pith_number":"pith:KDB5WWRX","schema_version":"1.0","canonical_sha256":"50c3db5a3739d4a60bcd04c9b5a7b5de408969b1885efbbf663e353eb1ee5b2c","source":{"kind":"arxiv","id":"2502.04370","version":1},"attestation_state":"computed","paper":{"title":"DreamDPO: Aligning Text-to-3D Generation with Human Preferences via Direct Preference Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.GR","cs.LG"],"primary_cat":"cs.CL","authors_text":"Fan Ma, Hehe Fan, Tat-Seng Chua, Xiaobo Xia, Yi Yang, Zhenglin Zhou","submitted_at":"2025-02-05T11:03:08Z","abstract_excerpt":"Text-to-3D generation automates 3D content creation from textual descriptions, which offers transformative potential across various fields. However, existing methods often struggle to align generated content with human preferences, limiting their applicability and flexibility. To address these limitations, in this paper, we propose DreamDPO, an optimization-based framework that integrates human preferences into the 3D generation process, through direct preference optimization. Practically, DreamDPO first constructs pairwise examples, then compare their alignment with human preferences using re"},"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":"2502.04370","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-05T11:03:08Z","cross_cats_sorted":["cs.GR","cs.LG"],"title_canon_sha256":"e91b837211072d5095058b4ab99f2586c9c9c713aa9b10d1511b80a42b0d4610","abstract_canon_sha256":"7c29d912aef10e3595bbbd378b824e23d2e6c81cd929e4d9006639a91156e9cd"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:10:50.494986Z","signature_b64":"FFvx0h5YwCmmKE+8Jombo5GAd4Ff+hxtl8WfWkcwajkGx9LkyDVSPwS259dxR272ewDkhmh4GjQqhgpNfNDQDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"50c3db5a3739d4a60bcd04c9b5a7b5de408969b1885efbbf663e353eb1ee5b2c","last_reissued_at":"2026-07-05T10:10:50.494517Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:10:50.494517Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"DreamDPO: Aligning Text-to-3D Generation with Human Preferences via Direct Preference Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.GR","cs.LG"],"primary_cat":"cs.CL","authors_text":"Fan Ma, Hehe Fan, Tat-Seng Chua, Xiaobo Xia, Yi Yang, Zhenglin Zhou","submitted_at":"2025-02-05T11:03:08Z","abstract_excerpt":"Text-to-3D generation automates 3D content creation from textual descriptions, which offers transformative potential across various fields. However, existing methods often struggle to align generated content with human preferences, limiting their applicability and flexibility. To address these limitations, in this paper, we propose DreamDPO, an optimization-based framework that integrates human preferences into the 3D generation process, through direct preference optimization. Practically, DreamDPO first constructs pairwise examples, then compare their alignment with human preferences using re"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.04370","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/2502.04370/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":"2502.04370","created_at":"2026-07-05T10:10:50.494576+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.04370v1","created_at":"2026-07-05T10:10:50.494576+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.04370","created_at":"2026-07-05T10:10:50.494576+00:00"},{"alias_kind":"pith_short_12","alias_value":"KDB5WWRXHHKK","created_at":"2026-07-05T10:10:50.494576+00:00"},{"alias_kind":"pith_short_16","alias_value":"KDB5WWRXHHKKMC6N","created_at":"2026-07-05T10:10:50.494576+00:00"},{"alias_kind":"pith_short_8","alias_value":"KDB5WWRX","created_at":"2026-07-05T10:10:50.494576+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.20364","citing_title":"Judging to Improve: A De-biased VLM-as-3D-Judge Protocol for Single-Image 3D Generation","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2606.18451","citing_title":"A Cross-Model VLM-Judge Protocol for Single-Image 3D Mesh Quality (and Why Cheap Proxies Fall Short)","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2605.26182","citing_title":"BrickAnything: Geometry-Conditioned Buildable Brick Generation with Structure-Aware Tokenization","ref_index":44,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/KDB5WWRXHHKKMC6NATE3LJ5V3Z","json":"https://pith.science/pith/KDB5WWRXHHKKMC6NATE3LJ5V3Z.json","graph_json":"https://pith.science/api/pith-number/KDB5WWRXHHKKMC6NATE3LJ5V3Z/graph.json","events_json":"https://pith.science/api/pith-number/KDB5WWRXHHKKMC6NATE3LJ5V3Z/events.json","paper":"https://pith.science/paper/KDB5WWRX"},"agent_actions":{"view_html":"https://pith.science/pith/KDB5WWRXHHKKMC6NATE3LJ5V3Z","download_json":"https://pith.science/pith/KDB5WWRXHHKKMC6NATE3LJ5V3Z.json","view_paper":"https://pith.science/paper/KDB5WWRX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.04370&json=true","fetch_graph":"https://pith.science/api/pith-number/KDB5WWRXHHKKMC6NATE3LJ5V3Z/graph.json","fetch_events":"https://pith.science/api/pith-number/KDB5WWRXHHKKMC6NATE3LJ5V3Z/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/KDB5WWRXHHKKMC6NATE3LJ5V3Z/action/timestamp_anchor","attest_storage":"https://pith.science/pith/KDB5WWRXHHKKMC6NATE3LJ5V3Z/action/storage_attestation","attest_author":"https://pith.science/pith/KDB5WWRXHHKKMC6NATE3LJ5V3Z/action/author_attestation","sign_citation":"https://pith.science/pith/KDB5WWRXHHKKMC6NATE3LJ5V3Z/action/citation_signature","submit_replication":"https://pith.science/pith/KDB5WWRXHHKKMC6NATE3LJ5V3Z/action/replication_record"}},"created_at":"2026-07-05T10:10:50.494576+00:00","updated_at":"2026-07-05T10:10:50.494576+00:00"}