{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:DVPSGXRBDDTEALQU5DVQORML44","short_pith_number":"pith:DVPSGXRB","schema_version":"1.0","canonical_sha256":"1d5f235e2118e6402e14e8eb07458be702e4326e8b630f83671de96cb3e42b6c","source":{"kind":"arxiv","id":"2207.04978","version":1},"attestation_state":"computed","paper":{"title":"Wave-ViT: Unifying Wavelet and Transformers for Visual Representation Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Chong-Wah Ngo, Tao Mei, Ting Yao, Yehao Li, Yingwei Pan","submitted_at":"2022-07-11T16:03:51Z","abstract_excerpt":"Multi-scale Vision Transformer (ViT) has emerged as a powerful backbone for computer vision tasks, while the self-attention computation in Transformer scales quadratically w.r.t. the input patch number. Thus, existing solutions commonly employ down-sampling operations (e.g., average pooling) over keys/values to dramatically reduce the computational cost. In this work, we argue that such over-aggressive down-sampling design is not invertible and inevitably causes information dropping especially for high-frequency components in objects (e.g., texture details). Motivated by the wavelet theory, we"},"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":"2207.04978","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2022-07-11T16:03:51Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"0894cb6369ad480231249c41a0e34ed684f3c1da9ac7dc02b44f7dc0220606bb","abstract_canon_sha256":"e6ca30caa93694bceb74200c26de4f221363c4309f01af8b97df4f2752b25c98"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:39:05.698728Z","signature_b64":"DvQrdeEdaoZDaJthG/8gDD0rEgYq9baV3hn7YhK7CmyYdjlpDEu/n+54yHa4xcsjZ1Xz+vH/2XpGQwGoeUmhAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1d5f235e2118e6402e14e8eb07458be702e4326e8b630f83671de96cb3e42b6c","last_reissued_at":"2026-07-05T04:39:05.697898Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:39:05.697898Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Wave-ViT: Unifying Wavelet and Transformers for Visual Representation Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Chong-Wah Ngo, Tao Mei, Ting Yao, Yehao Li, Yingwei Pan","submitted_at":"2022-07-11T16:03:51Z","abstract_excerpt":"Multi-scale Vision Transformer (ViT) has emerged as a powerful backbone for computer vision tasks, while the self-attention computation in Transformer scales quadratically w.r.t. the input patch number. Thus, existing solutions commonly employ down-sampling operations (e.g., average pooling) over keys/values to dramatically reduce the computational cost. In this work, we argue that such over-aggressive down-sampling design is not invertible and inevitably causes information dropping especially for high-frequency components in objects (e.g., texture details). Motivated by the wavelet theory, we"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2207.04978","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/2207.04978/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":"2207.04978","created_at":"2026-07-05T04:39:05.697967+00:00"},{"alias_kind":"arxiv_version","alias_value":"2207.04978v1","created_at":"2026-07-05T04:39:05.697967+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2207.04978","created_at":"2026-07-05T04:39:05.697967+00:00"},{"alias_kind":"pith_short_12","alias_value":"DVPSGXRBDDTE","created_at":"2026-07-05T04:39:05.697967+00:00"},{"alias_kind":"pith_short_16","alias_value":"DVPSGXRBDDTEALQU","created_at":"2026-07-05T04:39:05.697967+00:00"},{"alias_kind":"pith_short_8","alias_value":"DVPSGXRB","created_at":"2026-07-05T04:39:05.697967+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.10378","citing_title":"FSS-Net: Frequency-Spatial Synergy Network with Wavelet Attention for Carotid Artery Ultrasound Segmentation","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12026","citing_title":"Spectral Vision Transformer for Efficient Tokenization with Limited Data","ref_index":21,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DVPSGXRBDDTEALQU5DVQORML44","json":"https://pith.science/pith/DVPSGXRBDDTEALQU5DVQORML44.json","graph_json":"https://pith.science/api/pith-number/DVPSGXRBDDTEALQU5DVQORML44/graph.json","events_json":"https://pith.science/api/pith-number/DVPSGXRBDDTEALQU5DVQORML44/events.json","paper":"https://pith.science/paper/DVPSGXRB"},"agent_actions":{"view_html":"https://pith.science/pith/DVPSGXRBDDTEALQU5DVQORML44","download_json":"https://pith.science/pith/DVPSGXRBDDTEALQU5DVQORML44.json","view_paper":"https://pith.science/paper/DVPSGXRB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2207.04978&json=true","fetch_graph":"https://pith.science/api/pith-number/DVPSGXRBDDTEALQU5DVQORML44/graph.json","fetch_events":"https://pith.science/api/pith-number/DVPSGXRBDDTEALQU5DVQORML44/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DVPSGXRBDDTEALQU5DVQORML44/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DVPSGXRBDDTEALQU5DVQORML44/action/storage_attestation","attest_author":"https://pith.science/pith/DVPSGXRBDDTEALQU5DVQORML44/action/author_attestation","sign_citation":"https://pith.science/pith/DVPSGXRBDDTEALQU5DVQORML44/action/citation_signature","submit_replication":"https://pith.science/pith/DVPSGXRBDDTEALQU5DVQORML44/action/replication_record"}},"created_at":"2026-07-05T04:39:05.697967+00:00","updated_at":"2026-07-05T04:39:05.697967+00:00"}