{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:TM4JUTPDFS352CESMA5AHNESXM","short_pith_number":"pith:TM4JUTPD","schema_version":"1.0","canonical_sha256":"9b389a4de32cb7dd0892603a03b492bb345816bfb80a32f12a408d0aca643ad7","source":{"kind":"arxiv","id":"2607.20238","version":1},"attestation_state":"computed","paper":{"title":"Not All Patches are Equal: Sampling Matters for Visible-Infrared Pre-Training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bin Deng, Junjie Zhu, Ke Yang, Puhong Duan, Qiangjuan Huang, Qiwei Ma, Shutao Li, Xudong Kang","submitted_at":"2026-07-22T14:59:02Z","abstract_excerpt":"Visible-infrared (VIS-IR) alignment is a key pre-training task for robust multi-sensor perception. Most existing methods use uniform patch-wise contrastive learning, but this can be unreliable in VIS-IR data because imaging-physics differences make some spatially paired regions inherently less comparable, and aligning them with equal strength hinders representation learning and downstream transfer. In this paper, we revisit VIS-IR pre-training from a sampling perspective and propose Importance-Aware Sampling (IAS), which adjusts training emphasis based on patch reliability. Specifically, IAS ("},"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":"2607.20238","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-07-22T14:59:02Z","cross_cats_sorted":[],"title_canon_sha256":"161a797f22d6051cc3d67e7ee81e1a0fefd8edb43bbb65cfb06a97145d9b16a8","abstract_canon_sha256":"cdb550f29615f3d028a9c6a25799206e7121ec55d96cb2030f335300dcfc6606"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-23T01:25:09.540770Z","signature_b64":"9300SPwwDk6EemwmDIVs7VEL334ZxZmnbT62TD6oTXCK8gttcsYQ0rCK5Jpkue/6QsBqZ6SO2HrvUm3JHtaxBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9b389a4de32cb7dd0892603a03b492bb345816bfb80a32f12a408d0aca643ad7","last_reissued_at":"2026-07-23T01:25:09.539961Z","signature_status":"signed_v1","first_computed_at":"2026-07-23T01:25:09.539961Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Not All Patches are Equal: Sampling Matters for Visible-Infrared Pre-Training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bin Deng, Junjie Zhu, Ke Yang, Puhong Duan, Qiangjuan Huang, Qiwei Ma, Shutao Li, Xudong Kang","submitted_at":"2026-07-22T14:59:02Z","abstract_excerpt":"Visible-infrared (VIS-IR) alignment is a key pre-training task for robust multi-sensor perception. Most existing methods use uniform patch-wise contrastive learning, but this can be unreliable in VIS-IR data because imaging-physics differences make some spatially paired regions inherently less comparable, and aligning them with equal strength hinders representation learning and downstream transfer. In this paper, we revisit VIS-IR pre-training from a sampling perspective and propose Importance-Aware Sampling (IAS), which adjusts training emphasis based on patch reliability. Specifically, IAS ("},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.20238","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/2607.20238/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":"2607.20238","created_at":"2026-07-23T01:25:09.540384+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.20238v1","created_at":"2026-07-23T01:25:09.540384+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.20238","created_at":"2026-07-23T01:25:09.540384+00:00"},{"alias_kind":"pith_short_12","alias_value":"TM4JUTPDFS35","created_at":"2026-07-23T01:25:09.540384+00:00"},{"alias_kind":"pith_short_16","alias_value":"TM4JUTPDFS352CES","created_at":"2026-07-23T01:25:09.540384+00:00"},{"alias_kind":"pith_short_8","alias_value":"TM4JUTPD","created_at":"2026-07-23T01:25:09.540384+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/TM4JUTPDFS352CESMA5AHNESXM","json":"https://pith.science/pith/TM4JUTPDFS352CESMA5AHNESXM.json","graph_json":"https://pith.science/api/pith-number/TM4JUTPDFS352CESMA5AHNESXM/graph.json","events_json":"https://pith.science/api/pith-number/TM4JUTPDFS352CESMA5AHNESXM/events.json","paper":"https://pith.science/paper/TM4JUTPD"},"agent_actions":{"view_html":"https://pith.science/pith/TM4JUTPDFS352CESMA5AHNESXM","download_json":"https://pith.science/pith/TM4JUTPDFS352CESMA5AHNESXM.json","view_paper":"https://pith.science/paper/TM4JUTPD","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.20238&json=true","fetch_graph":"https://pith.science/api/pith-number/TM4JUTPDFS352CESMA5AHNESXM/graph.json","fetch_events":"https://pith.science/api/pith-number/TM4JUTPDFS352CESMA5AHNESXM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TM4JUTPDFS352CESMA5AHNESXM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TM4JUTPDFS352CESMA5AHNESXM/action/storage_attestation","attest_author":"https://pith.science/pith/TM4JUTPDFS352CESMA5AHNESXM/action/author_attestation","sign_citation":"https://pith.science/pith/TM4JUTPDFS352CESMA5AHNESXM/action/citation_signature","submit_replication":"https://pith.science/pith/TM4JUTPDFS352CESMA5AHNESXM/action/replication_record"}},"created_at":"2026-07-23T01:25:09.540384+00:00","updated_at":"2026-07-23T01:25:09.540384+00:00"}