{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:2TC26XF5URH4UIICN23M7DF5XM","short_pith_number":"pith:2TC26XF5","schema_version":"1.0","canonical_sha256":"d4c5af5cbda44fca21026eb6cf8cbdbb1335abcb38f82d0c93635a799518ab6e","source":{"kind":"arxiv","id":"2106.09011","version":2},"attestation_state":"computed","paper":{"title":"Evolving Image Compositions for Feature Representation Learning","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.LG","cs.NE"],"primary_cat":"cs.CV","authors_text":"Arshdeep Sekhon, Paola Cascante-Bonilla, Vicente Ordonez, Yanjun Qi","submitted_at":"2021-06-16T17:57:18Z","abstract_excerpt":"Convolutional neural networks for visual recognition require large amounts of training samples and usually benefit from data augmentation. This paper proposes PatchMix, a data augmentation method that creates new samples by composing patches from pairs of images in a grid-like pattern. These new samples are assigned label scores that are proportional to the number of patches borrowed from each image. We then add a set of additional losses at the patch-level to regularize and to encourage good representations at both the patch and image levels. A ResNet-50 model trained on ImageNet using PatchM"},"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":"2106.09011","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2021-06-16T17:57:18Z","cross_cats_sorted":["cs.LG","cs.NE"],"title_canon_sha256":"cdc1b3f82b4440fbdc49a50ec4a47c7b4ab28bf0154aef6882d2903c511ca56a","abstract_canon_sha256":"8ed6e8a7a9328ac792416c9a9f55c797c634fac878ce73e37c1e0923095ea814"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:10:36.000665Z","signature_b64":"wCwLmpWTP3XXFCoI7o/dAfQU2XZxuSndDCSygsMsTnCpwyqfMcxc/4uf8PpFki6MQKQx9jSpUykMv7tHaertDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d4c5af5cbda44fca21026eb6cf8cbdbb1335abcb38f82d0c93635a799518ab6e","last_reissued_at":"2026-07-05T04:10:36.000189Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:10:36.000189Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Evolving Image Compositions for Feature Representation Learning","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.LG","cs.NE"],"primary_cat":"cs.CV","authors_text":"Arshdeep Sekhon, Paola Cascante-Bonilla, Vicente Ordonez, Yanjun Qi","submitted_at":"2021-06-16T17:57:18Z","abstract_excerpt":"Convolutional neural networks for visual recognition require large amounts of training samples and usually benefit from data augmentation. This paper proposes PatchMix, a data augmentation method that creates new samples by composing patches from pairs of images in a grid-like pattern. These new samples are assigned label scores that are proportional to the number of patches borrowed from each image. We then add a set of additional losses at the patch-level to regularize and to encourage good representations at both the patch and image levels. A ResNet-50 model trained on ImageNet using PatchM"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2106.09011","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/2106.09011/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":"2106.09011","created_at":"2026-07-05T04:10:36.000246+00:00"},{"alias_kind":"arxiv_version","alias_value":"2106.09011v2","created_at":"2026-07-05T04:10:36.000246+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2106.09011","created_at":"2026-07-05T04:10:36.000246+00:00"},{"alias_kind":"pith_short_12","alias_value":"2TC26XF5URH4","created_at":"2026-07-05T04:10:36.000246+00:00"},{"alias_kind":"pith_short_16","alias_value":"2TC26XF5URH4UIIC","created_at":"2026-07-05T04:10:36.000246+00:00"},{"alias_kind":"pith_short_8","alias_value":"2TC26XF5","created_at":"2026-07-05T04:10:36.000246+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/2TC26XF5URH4UIICN23M7DF5XM","json":"https://pith.science/pith/2TC26XF5URH4UIICN23M7DF5XM.json","graph_json":"https://pith.science/api/pith-number/2TC26XF5URH4UIICN23M7DF5XM/graph.json","events_json":"https://pith.science/api/pith-number/2TC26XF5URH4UIICN23M7DF5XM/events.json","paper":"https://pith.science/paper/2TC26XF5"},"agent_actions":{"view_html":"https://pith.science/pith/2TC26XF5URH4UIICN23M7DF5XM","download_json":"https://pith.science/pith/2TC26XF5URH4UIICN23M7DF5XM.json","view_paper":"https://pith.science/paper/2TC26XF5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2106.09011&json=true","fetch_graph":"https://pith.science/api/pith-number/2TC26XF5URH4UIICN23M7DF5XM/graph.json","fetch_events":"https://pith.science/api/pith-number/2TC26XF5URH4UIICN23M7DF5XM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2TC26XF5URH4UIICN23M7DF5XM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2TC26XF5URH4UIICN23M7DF5XM/action/storage_attestation","attest_author":"https://pith.science/pith/2TC26XF5URH4UIICN23M7DF5XM/action/author_attestation","sign_citation":"https://pith.science/pith/2TC26XF5URH4UIICN23M7DF5XM/action/citation_signature","submit_replication":"https://pith.science/pith/2TC26XF5URH4UIICN23M7DF5XM/action/replication_record"}},"created_at":"2026-07-05T04:10:36.000246+00:00","updated_at":"2026-07-05T04:10:36.000246+00:00"}