{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:HCVTMNFBYQPIBVEUTLUYHCYHQG","short_pith_number":"pith:HCVTMNFB","schema_version":"1.0","canonical_sha256":"38ab3634a1c41e80d4949ae9838b078198a2057ce69edd9fc586955d13396723","source":{"kind":"arxiv","id":"2007.13003","version":3},"attestation_state":"computed","paper":{"title":"Robust and Generalizable Visual Representation Learning via Random Convolutions","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Colin Raffel, Deyi Liu, Junlin Yang, Marc Niethammer, Zhenlin Xu","submitted_at":"2020-07-25T19:52:25Z","abstract_excerpt":"While successful for various computer vision tasks, deep neural networks have shown to be vulnerable to texture style shifts and small perturbations to which humans are robust. In this work, we show that the robustness of neural networks can be greatly improved through the use of random convolutions as data augmentation. Random convolutions are approximately shape-preserving and may distort local textures. Intuitively, randomized convolutions create an infinite number of new domains with similar global shapes but random local textures. Therefore, we explore using outputs of multi-scale random "},"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":"2007.13003","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2020-07-25T19:52:25Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"9613ebd002da8fe8332cbdb557d8691d0509e72de52fd3f7cee3fd124cce32d6","abstract_canon_sha256":"f387a8ed9607b066ff68f368c1d8cfe86231fade77fb05704478642a851dc052"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:36:44.746730Z","signature_b64":"tQK56//ZGX90hKMsCOm7EyGBaYSrB16KJE4y7rv7eWlfj0JkqHyniBQitMJUQ+9qzxuyBYSiE2/BTAUPY70yBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"38ab3634a1c41e80d4949ae9838b078198a2057ce69edd9fc586955d13396723","last_reissued_at":"2026-07-05T02:36:44.746167Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:36:44.746167Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Robust and Generalizable Visual Representation Learning via Random Convolutions","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Colin Raffel, Deyi Liu, Junlin Yang, Marc Niethammer, Zhenlin Xu","submitted_at":"2020-07-25T19:52:25Z","abstract_excerpt":"While successful for various computer vision tasks, deep neural networks have shown to be vulnerable to texture style shifts and small perturbations to which humans are robust. In this work, we show that the robustness of neural networks can be greatly improved through the use of random convolutions as data augmentation. Random convolutions are approximately shape-preserving and may distort local textures. Intuitively, randomized convolutions create an infinite number of new domains with similar global shapes but random local textures. Therefore, we explore using outputs of multi-scale random "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2007.13003","kind":"arxiv","version":3},"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/2007.13003/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":"2007.13003","created_at":"2026-07-05T02:36:44.746233+00:00"},{"alias_kind":"arxiv_version","alias_value":"2007.13003v3","created_at":"2026-07-05T02:36:44.746233+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2007.13003","created_at":"2026-07-05T02:36:44.746233+00:00"},{"alias_kind":"pith_short_12","alias_value":"HCVTMNFBYQPI","created_at":"2026-07-05T02:36:44.746233+00:00"},{"alias_kind":"pith_short_16","alias_value":"HCVTMNFBYQPIBVEU","created_at":"2026-07-05T02:36:44.746233+00:00"},{"alias_kind":"pith_short_8","alias_value":"HCVTMNFB","created_at":"2026-07-05T02:36:44.746233+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.19649","citing_title":"CAD-Free Learning of Spacecraft Pose Estimators via NeRF-Based Augmentations","ref_index":39,"is_internal_anchor":false},{"citing_arxiv_id":"2605.19649","citing_title":"CAD-Free Learning of Spacecraft Pose Estimators via NeRF-Based Augmentations","ref_index":39,"is_internal_anchor":false},{"citing_arxiv_id":"2604.02564","citing_title":"Why Invariance is Not Enough for Biomedical Domain Generalization and How to Fix It","ref_index":65,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09925","citing_title":"Frequency Adapter with SAM for Generalized Medical Image Segmentation","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2604.10524","citing_title":"FGML-DG: Feynman-Inspired Cognitive Science Paradigm for Cross-Domain Medical Image Segmentation","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2605.03098","citing_title":"One Sequence to Segment Them All: Efficient Data Augmentation for CT and MRI Cross-Domain 3D Spine Segmentation","ref_index":23,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/HCVTMNFBYQPIBVEUTLUYHCYHQG","json":"https://pith.science/pith/HCVTMNFBYQPIBVEUTLUYHCYHQG.json","graph_json":"https://pith.science/api/pith-number/HCVTMNFBYQPIBVEUTLUYHCYHQG/graph.json","events_json":"https://pith.science/api/pith-number/HCVTMNFBYQPIBVEUTLUYHCYHQG/events.json","paper":"https://pith.science/paper/HCVTMNFB"},"agent_actions":{"view_html":"https://pith.science/pith/HCVTMNFBYQPIBVEUTLUYHCYHQG","download_json":"https://pith.science/pith/HCVTMNFBYQPIBVEUTLUYHCYHQG.json","view_paper":"https://pith.science/paper/HCVTMNFB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2007.13003&json=true","fetch_graph":"https://pith.science/api/pith-number/HCVTMNFBYQPIBVEUTLUYHCYHQG/graph.json","fetch_events":"https://pith.science/api/pith-number/HCVTMNFBYQPIBVEUTLUYHCYHQG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HCVTMNFBYQPIBVEUTLUYHCYHQG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HCVTMNFBYQPIBVEUTLUYHCYHQG/action/storage_attestation","attest_author":"https://pith.science/pith/HCVTMNFBYQPIBVEUTLUYHCYHQG/action/author_attestation","sign_citation":"https://pith.science/pith/HCVTMNFBYQPIBVEUTLUYHCYHQG/action/citation_signature","submit_replication":"https://pith.science/pith/HCVTMNFBYQPIBVEUTLUYHCYHQG/action/replication_record"}},"created_at":"2026-07-05T02:36:44.746233+00:00","updated_at":"2026-07-05T02:36:44.746233+00:00"}