{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:6NIQKDSFOSW3JHRJF32XEJI5US","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"09cd875eb0e73572935920e3e662d2ea29577bdd4cac5262791b25635e48cad7","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-04-07T10:40:27Z","title_canon_sha256":"cc48b7e54d393b0777d6f02aa9d23aa590570315da226405be64278dd018b994"},"schema_version":"1.0","source":{"id":"2604.05689","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2604.05689","created_at":"2026-06-30T02:18:08Z"},{"alias_kind":"arxiv_version","alias_value":"2604.05689v1","created_at":"2026-06-30T02:18:08Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2604.05689","created_at":"2026-06-30T02:18:08Z"},{"alias_kind":"pith_short_12","alias_value":"6NIQKDSFOSW3","created_at":"2026-06-30T02:18:08Z"},{"alias_kind":"pith_short_16","alias_value":"6NIQKDSFOSW3JHRJ","created_at":"2026-06-30T02:18:08Z"},{"alias_kind":"pith_short_8","alias_value":"6NIQKDSF","created_at":"2026-06-30T02:18:08Z"}],"graph_snapshots":[{"event_id":"sha256:a6912700c56243446d18bf985c1cb68eb19d528e94fb5dd1fe0a0bd5cadeea05","target":"graph","created_at":"2026-06-30T02:18:08Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"CRFT consistently outperforms state-of-the-art registration methods in both accuracy and robustness."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That a single modality-independent feature flow representation learned in a transformer can jointly handle feature alignment and flow estimation while the iterative discrepancy-guided attention with Spatial Geometric Transform enforces consistency under large affine and scale variations."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"CRFT is a new transformer architecture using recurrent consistent feature flow learning to achieve accurate and robust cross-modal image registration under large variations."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"CRFT uses a transformer to learn a consistent recurrent feature flow that aligns cross-modal images more accurately and robustly than existing methods."}],"snapshot_sha256":"042c5fcdb45fe935cdcb410e49d6c1fa51c1e4e6d64cf100da03bbdbb48c5a1c"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"424e65ad8df31fda6f16ad8c9beede6a169ea1514d1e782905ad5cab8f1f8ae5"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2604.05689/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We present Consistent-Recurrent Feature Flow Transformer (CRFT), a unified coarse-to-fine framework based on feature flow learning for robust cross-modal image registration. CRFT learns a modality-independent feature flow representation within a transformer-based architecture that jointly performs feature alignment and flow estimation. The coarse stage establishes global correspondences through multi-scale feature correlation, while the fine stage refines local details via hierarchical feature fusion and adaptive spatial reasoning. To enhance geometric adaptability, an iterative discrepancy-gu","authors_text":"Mengzhu Ding, Xichao Teng, Xuecong Liu, Zhang Li, Zixuan Sun","cross_cats":["cs.AI"],"headline":"CRFT uses a transformer to learn a consistent recurrent feature flow that aligns cross-modal images more accurately and robustly than existing methods.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-04-07T10:40:27Z","title":"CRFT: Consistent-Recurrent Feature Flow Transformer for Cross-Modal Image Registration"},"references":{"count":66,"internal_anchors":0,"resolved_work":66,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"Deep learning models for digital image processing: a review.Artificial Intelligence Review, 57(1):11","work_id":"10f36277-2b3b-4126-8c6e-27acf1e62bfc","year":2024},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"Graphi2p: Image-to-point cloud registration with exploring pattern of correspondence via graph learning","work_id":"f9c7a4d2-7191-4470-bf62-3c05728a2b15","year":2025},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"A survey on deep learning in medical image registration: New technologies, uncertainty, evaluation metrics, and beyond.Medical Image Analysis, 100:103385","work_id":"393cbdbe-f3a8-4461-a62b-14adea99f0be","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"Mvsplat: Efficient 3d gaussian splatting from sparse multi-view images","work_id":"7e9dc93b-bac5-4ee5-8ee6-9917a40c6c06","year":2024},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"Dsap: Dynamic sparse attention perception matcher for accurate local feature matching","work_id":"3fe403b5-98e3-425e-b483-a3fe94790fa8","year":2024}],"snapshot_sha256":"6f6fd9388ae90e1b7e52e7aceef8ccf80ada3a918399797e2c585de13637f91d"},"source":{"id":"2604.05689","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-10T19:34:16.218576Z","id":"49fafc63-3361-4266-af2e-93324b8ef09a","model_set":{"reader":"grok-4.3"},"one_line_summary":"CRFT is a new transformer architecture using recurrent consistent feature flow learning to achieve accurate and robust cross-modal image registration under large variations.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"CRFT uses a transformer to learn a consistent recurrent feature flow that aligns cross-modal images more accurately and robustly than existing methods.","strongest_claim":"CRFT consistently outperforms state-of-the-art registration methods in both accuracy and robustness.","weakest_assumption":"That a single modality-independent feature flow representation learned in a transformer can jointly handle feature alignment and flow estimation while the iterative discrepancy-guided attention with Spatial Geometric Transform enforces consistency under large affine and scale variations."}},"verdict_id":"49fafc63-3361-4266-af2e-93324b8ef09a"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:7d4e2ad7d90e5bfd8e4e78690983d00bb39bf34a605bd7ec9190bfadfd422f72","target":"record","created_at":"2026-06-30T02:18:08Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"09cd875eb0e73572935920e3e662d2ea29577bdd4cac5262791b25635e48cad7","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-04-07T10:40:27Z","title_canon_sha256":"cc48b7e54d393b0777d6f02aa9d23aa590570315da226405be64278dd018b994"},"schema_version":"1.0","source":{"id":"2604.05689","kind":"arxiv","version":1}},"canonical_sha256":"f351050e4574adb49e292ef572251da4bb3724265d4f0ab70d8e158c903360df","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f351050e4574adb49e292ef572251da4bb3724265d4f0ab70d8e158c903360df","first_computed_at":"2026-06-30T02:18:08.151128Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-30T02:18:08.151128Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"yp8wdcS0qPzg+SybfDul890Bkfm7cYSnBpV4DGy80XLycpVC+AW9W6AsP7iCKL8eNF1pMAVa60AnwQkhCcqPBA==","signature_status":"signed_v1","signed_at":"2026-06-30T02:18:08.151802Z","signed_message":"canonical_sha256_bytes"},"source_id":"2604.05689","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:7d4e2ad7d90e5bfd8e4e78690983d00bb39bf34a605bd7ec9190bfadfd422f72","sha256:a6912700c56243446d18bf985c1cb68eb19d528e94fb5dd1fe0a0bd5cadeea05"],"state_sha256":"203c671946d409eff1ec5b547fe4b7b0c0f648b5f2816af40cc8b3dc49ecb46c"}