{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:FDOMKKJUFSYMGAZCZNOM76SQ7Q","short_pith_number":"pith:FDOMKKJU","schema_version":"1.0","canonical_sha256":"28dcc529342cb0c30322cb5ccffa50fc2f4fec423a7fee396fcaf5bd4ef92df3","source":{"kind":"arxiv","id":"2311.17138","version":2},"attestation_state":"computed","paper":{"title":"Shadows Don't Lie and Lines Can't Bend! Generative Models don't know Projective Geometry...for now","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.GR","cs.LG"],"primary_cat":"cs.CV","authors_text":"Amitabh Mahapatra, Anand Bhattad, Ayush Sarkar, D.A. Forsyth, Hanlin Mai, Svetlana Lazebnik","submitted_at":"2023-11-28T18:59:06Z","abstract_excerpt":"Generative models can produce impressively realistic images. This paper demonstrates that generated images have geometric features different from those of real images. We build a set of collections of generated images, prequalified to fool simple, signal-based classifiers into believing they are real. We then show that prequalified generated images can be identified reliably by classifiers that only look at geometric properties. We use three such classifiers. All three classifiers are denied access to image pixels, and look only at derived geometric features. The first classifier looks at the "},"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":"2311.17138","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2023-11-28T18:59:06Z","cross_cats_sorted":["cs.AI","cs.GR","cs.LG"],"title_canon_sha256":"42b8142bd41fe3533c69c2ec3da9c7baded75aea8eca29b62d0200536bfb096e","abstract_canon_sha256":"2725d038240a5965f52ef3e68ddd2f679bf67f5452fae7fba4d3f26009fac5ba"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:25:30.237395Z","signature_b64":"76XjOmd23k57lD3bDUvd5SnPEXdCcsvTysCqfyyDoTag8JcG0MR1AIiHhtCDceangd4MyFeXDCD2En4dFZlFCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"28dcc529342cb0c30322cb5ccffa50fc2f4fec423a7fee396fcaf5bd4ef92df3","last_reissued_at":"2026-07-05T08:25:30.236954Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:25:30.236954Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Shadows Don't Lie and Lines Can't Bend! Generative Models don't know Projective Geometry...for now","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.GR","cs.LG"],"primary_cat":"cs.CV","authors_text":"Amitabh Mahapatra, Anand Bhattad, Ayush Sarkar, D.A. Forsyth, Hanlin Mai, Svetlana Lazebnik","submitted_at":"2023-11-28T18:59:06Z","abstract_excerpt":"Generative models can produce impressively realistic images. This paper demonstrates that generated images have geometric features different from those of real images. We build a set of collections of generated images, prequalified to fool simple, signal-based classifiers into believing they are real. We then show that prequalified generated images can be identified reliably by classifiers that only look at geometric properties. We use three such classifiers. All three classifiers are denied access to image pixels, and look only at derived geometric features. The first classifier looks at the "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2311.17138","kind":"arxiv","version":2},"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/2311.17138/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":"2311.17138","created_at":"2026-07-05T08:25:30.237008+00:00"},{"alias_kind":"arxiv_version","alias_value":"2311.17138v2","created_at":"2026-07-05T08:25:30.237008+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2311.17138","created_at":"2026-07-05T08:25:30.237008+00:00"},{"alias_kind":"pith_short_12","alias_value":"FDOMKKJUFSYM","created_at":"2026-07-05T08:25:30.237008+00:00"},{"alias_kind":"pith_short_16","alias_value":"FDOMKKJUFSYMGAZC","created_at":"2026-07-05T08:25:30.237008+00:00"},{"alias_kind":"pith_short_8","alias_value":"FDOMKKJU","created_at":"2026-07-05T08:25:30.237008+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2405.10314","citing_title":"CAT3D: Create Anything in 3D with Multi-View Diffusion Models","ref_index":70,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/FDOMKKJUFSYMGAZCZNOM76SQ7Q","json":"https://pith.science/pith/FDOMKKJUFSYMGAZCZNOM76SQ7Q.json","graph_json":"https://pith.science/api/pith-number/FDOMKKJUFSYMGAZCZNOM76SQ7Q/graph.json","events_json":"https://pith.science/api/pith-number/FDOMKKJUFSYMGAZCZNOM76SQ7Q/events.json","paper":"https://pith.science/paper/FDOMKKJU"},"agent_actions":{"view_html":"https://pith.science/pith/FDOMKKJUFSYMGAZCZNOM76SQ7Q","download_json":"https://pith.science/pith/FDOMKKJUFSYMGAZCZNOM76SQ7Q.json","view_paper":"https://pith.science/paper/FDOMKKJU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2311.17138&json=true","fetch_graph":"https://pith.science/api/pith-number/FDOMKKJUFSYMGAZCZNOM76SQ7Q/graph.json","fetch_events":"https://pith.science/api/pith-number/FDOMKKJUFSYMGAZCZNOM76SQ7Q/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FDOMKKJUFSYMGAZCZNOM76SQ7Q/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FDOMKKJUFSYMGAZCZNOM76SQ7Q/action/storage_attestation","attest_author":"https://pith.science/pith/FDOMKKJUFSYMGAZCZNOM76SQ7Q/action/author_attestation","sign_citation":"https://pith.science/pith/FDOMKKJUFSYMGAZCZNOM76SQ7Q/action/citation_signature","submit_replication":"https://pith.science/pith/FDOMKKJUFSYMGAZCZNOM76SQ7Q/action/replication_record"}},"created_at":"2026-07-05T08:25:30.237008+00:00","updated_at":"2026-07-05T08:25:30.237008+00:00"}