{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:4JSU4GTYLBM5ANJKP7ASLID2ZI","short_pith_number":"pith:4JSU4GTY","schema_version":"1.0","canonical_sha256":"e2654e1a785859d0352a7fc125a07aca2671027e1ffbb51076e76e9b69449adc","source":{"kind":"arxiv","id":"2208.11356","version":2},"attestation_state":"computed","paper":{"title":"Towards Efficient Use of Multi-Scale Features in Transformer-Based Object Detectors","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.LG","cs.MM"],"primary_cat":"cs.CV","authors_text":"Gongjie Zhang, Jingyi Zhang, Shijian Lu, Xiaoqin Zhang, Zhipeng Luo, Zichen Tian","submitted_at":"2022-08-24T08:09:25Z","abstract_excerpt":"Multi-scale features have been proven highly effective for object detection but often come with huge and even prohibitive extra computation costs, especially for the recent Transformer-based detectors. In this paper, we propose Iterative Multi-scale Feature Aggregation (IMFA) -- a generic paradigm that enables efficient use of multi-scale features in Transformer-based object detectors. The core idea is to exploit sparse multi-scale features from just a few crucial locations, and it is achieved with two novel designs. First, IMFA rearranges the Transformer encoder-decoder pipeline so that the e"},"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":"2208.11356","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2022-08-24T08:09:25Z","cross_cats_sorted":["cs.AI","cs.LG","cs.MM"],"title_canon_sha256":"ea82a64557d727b08a110dbefb8b1c8ed328d5c4f2a85a5455f9e8bf54e913e9","abstract_canon_sha256":"3e47430c386f2457fe9d7daab88ca605fd1748312f65052b7e420f15d205598f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:54:12.583928Z","signature_b64":"xQPboic4HtISn+ddCZnKj4NFaGuZhPDUZ4yLG1nTh7ZZumq5a8fefK3igzjJh0BD2ra+aAFX276lIB/6v6EnBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e2654e1a785859d0352a7fc125a07aca2671027e1ffbb51076e76e9b69449adc","last_reissued_at":"2026-07-05T05:54:12.583479Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:54:12.583479Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Towards Efficient Use of Multi-Scale Features in Transformer-Based Object Detectors","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.LG","cs.MM"],"primary_cat":"cs.CV","authors_text":"Gongjie Zhang, Jingyi Zhang, Shijian Lu, Xiaoqin Zhang, Zhipeng Luo, Zichen Tian","submitted_at":"2022-08-24T08:09:25Z","abstract_excerpt":"Multi-scale features have been proven highly effective for object detection but often come with huge and even prohibitive extra computation costs, especially for the recent Transformer-based detectors. In this paper, we propose Iterative Multi-scale Feature Aggregation (IMFA) -- a generic paradigm that enables efficient use of multi-scale features in Transformer-based object detectors. The core idea is to exploit sparse multi-scale features from just a few crucial locations, and it is achieved with two novel designs. First, IMFA rearranges the Transformer encoder-decoder pipeline so that the e"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2208.11356","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/2208.11356/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":"2208.11356","created_at":"2026-07-05T05:54:12.583537+00:00"},{"alias_kind":"arxiv_version","alias_value":"2208.11356v2","created_at":"2026-07-05T05:54:12.583537+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2208.11356","created_at":"2026-07-05T05:54:12.583537+00:00"},{"alias_kind":"pith_short_12","alias_value":"4JSU4GTYLBM5","created_at":"2026-07-05T05:54:12.583537+00:00"},{"alias_kind":"pith_short_16","alias_value":"4JSU4GTYLBM5ANJK","created_at":"2026-07-05T05:54:12.583537+00:00"},{"alias_kind":"pith_short_8","alias_value":"4JSU4GTY","created_at":"2026-07-05T05:54:12.583537+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/4JSU4GTYLBM5ANJKP7ASLID2ZI","json":"https://pith.science/pith/4JSU4GTYLBM5ANJKP7ASLID2ZI.json","graph_json":"https://pith.science/api/pith-number/4JSU4GTYLBM5ANJKP7ASLID2ZI/graph.json","events_json":"https://pith.science/api/pith-number/4JSU4GTYLBM5ANJKP7ASLID2ZI/events.json","paper":"https://pith.science/paper/4JSU4GTY"},"agent_actions":{"view_html":"https://pith.science/pith/4JSU4GTYLBM5ANJKP7ASLID2ZI","download_json":"https://pith.science/pith/4JSU4GTYLBM5ANJKP7ASLID2ZI.json","view_paper":"https://pith.science/paper/4JSU4GTY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2208.11356&json=true","fetch_graph":"https://pith.science/api/pith-number/4JSU4GTYLBM5ANJKP7ASLID2ZI/graph.json","fetch_events":"https://pith.science/api/pith-number/4JSU4GTYLBM5ANJKP7ASLID2ZI/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4JSU4GTYLBM5ANJKP7ASLID2ZI/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4JSU4GTYLBM5ANJKP7ASLID2ZI/action/storage_attestation","attest_author":"https://pith.science/pith/4JSU4GTYLBM5ANJKP7ASLID2ZI/action/author_attestation","sign_citation":"https://pith.science/pith/4JSU4GTYLBM5ANJKP7ASLID2ZI/action/citation_signature","submit_replication":"https://pith.science/pith/4JSU4GTYLBM5ANJKP7ASLID2ZI/action/replication_record"}},"created_at":"2026-07-05T05:54:12.583537+00:00","updated_at":"2026-07-05T05:54:12.583537+00:00"}