{"paper":{"title":"P2GS: Physical Prior-guided Gaussian Splatting for Photometrically Consistent Urban Reconstruction","license":"http://creativecommons.org/licenses/by/4.0/","headline":"P2GS jointly decomposes a view-invariant linear HDR radiance field, per-view exposure scales, and tone-mapping functions from LDR images to fix photometric inconsistencies in 3D Gaussian Splatting for urban scenes.","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Hidehisa Arai, Hironobu Fujiyoshi, Kota Shimomura, Takayoshi Yamashita, Tsubasa Takahashi","submitted_at":"2026-05-16T10:36:23Z","abstract_excerpt":"3D Gaussian Splatting (3DGS) has recently emerged as a powerful explicit representation enabling fast, high-fidelity rendering, making it a promising foundation for closed-loop simulators and perception models in autonomous driving. However, conventional 3DGS implicitly assumes consistent exposure and tone mapping across views. Real driving data violates this assumption due to heterogeneous camera pipelines and dynamic outdoor illumination, baking exposure discrepancies and sensor noise into the radiance field and producing artifacts and inconsistent illumination especially in static backgroun"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"P2GS jointly decomposes a view-invariant linear HDR radiance field, per-view exposure scales, and tone-mapping functions from only LDR images without HDR supervision, yielding a radiance field robust to inter-camera illumination differences while preserving real-time efficiency.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The physical image-formation process can be accurately inverted by jointly optimizing a shared linear HDR radiance field together with per-view exposure scales and tone-mapping functions from LDR observations alone, with relative-exposure consistency and HDR-domain regularization sufficient to prevent degenerate solutions.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"P2GS jointly decomposes LDR images into a view-invariant linear HDR radiance field, per-view exposure scales, and tone-mapping functions without HDR supervision to enforce photometric consistency in urban Gaussian Splatting.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"P2GS jointly decomposes a view-invariant linear HDR radiance field, per-view exposure scales, and tone-mapping functions from LDR images to fix photometric inconsistencies in 3D Gaussian Splatting for urban scenes.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"bdf338ee64a240a639c90aabcc97e6da9066637a6d1966970fc55a44476b5f14"},"source":{"id":"2605.16925","kind":"arxiv","version":1},"verdict":{"id":"98a1f36e-3f3f-4a75-86eb-71a92425771a","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-19T20:54:31.149506Z","strongest_claim":"P2GS jointly decomposes a view-invariant linear HDR radiance field, per-view exposure scales, and tone-mapping functions from only LDR images without HDR supervision, yielding a radiance field robust to inter-camera illumination differences while preserving real-time efficiency.","one_line_summary":"P2GS jointly decomposes LDR images into a view-invariant linear HDR radiance field, per-view exposure scales, and tone-mapping functions without HDR supervision to enforce photometric consistency in urban Gaussian Splatting.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The physical image-formation process can be accurately inverted by jointly optimizing a shared linear HDR radiance field together with per-view exposure scales and tone-mapping functions from LDR observations alone, with relative-exposure consistency and HDR-domain regularization sufficient to prevent degenerate solutions.","pith_extraction_headline":"P2GS jointly decomposes a view-invariant linear HDR radiance field, per-view exposure scales, and tone-mapping functions from LDR images to fix photometric inconsistencies in 3D Gaussian Splatting for urban scenes."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.16925/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"doi_title_agreement","ran_at":"2026-05-19T21:01:19.160367Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T21:00:55.203928Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"cited_work_retraction","ran_at":"2026-05-19T20:22:26.750397Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"claim_evidence","ran_at":"2026-05-19T18:41:56.258807Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T18:33:26.339608Z","status":"skipped","version":"1.0.0","findings_count":0}],"snapshot_sha256":"84bcd71f90619e555d82cb0e912d10c4b9a64cbb9f3e0b8eee1d5eb876cdb6f3"},"references":{"count":28,"sample":[{"doi":"","year":2024,"title":"Hdr-gs: Efficient high dynamic range novel view synthesis at 1000x speed via gaussian splatting","work_id":"8c657724-adc8-43fa-9303-be621b406113","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"Pseudo- simulation for autonomous driving","work_id":"d5bd3cd5-08f6-416f-96e5-9fe185d3aee1","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2022,"title":"Hallucinated neural radi- ance fields in the wild","work_id":"138cf31c-0047-4a26-9d1f-71297937bee3","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Periodic vibration gaussian: Dynamic urban scene reconstruction and real-time rendering","work_id":"b2ef8501-a724-4c23-9a5c-eb58122910f8","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"Omnire: Omni urban scene reconstruction","work_id":"bbec8a27-4b8d-41c8-8da9-004903f7ad77","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":28,"snapshot_sha256":"e771c48f0b9cb4e3c77480cb8c09fc9de5399ea957415ff1a839c5168938c8b0","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"}