{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2016:ETM2SHTA43LD7U4QT57OC7ATGV","short_pith_number":"pith:ETM2SHTA","schema_version":"1.0","canonical_sha256":"24d9a91e60e6d63fd3909f7ee17c1335484acb9f4b5e058a146569e7f67bd5b6","source":{"kind":"arxiv","id":"1608.00507","version":1},"attestation_state":"computed","paper":{"title":"Top-down Neural Attention by Excitation Backprop","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Jianming Zhang, Jonathan Brandt, Stan Sclaroff, Xiaohui Shen, Zhe Lin","submitted_at":"2016-08-01T17:49:57Z","abstract_excerpt":"We aim to model the top-down attention of a Convolutional Neural Network (CNN) classifier for generating task-specific attention maps. Inspired by a top-down human visual attention model, we propose a new backpropagation scheme, called Excitation Backprop, to pass along top-down signals downwards in the network hierarchy via a probabilistic Winner-Take-All process. Furthermore, we introduce the concept of contrastive attention to make the top-down attention maps more discriminative. In experiments, we demonstrate the accuracy and generalizability of our method in weakly supervised localization"},"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":"1608.00507","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2016-08-01T17:49:57Z","cross_cats_sorted":[],"title_canon_sha256":"439c0b8b33113690e02a9ccd6cab77c85723aad047678658db58066df9807405","abstract_canon_sha256":"2f4df305a2645f4f062cfa07db2df724819c36b2eb2fb7ab2a1459c824658c97"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T01:10:11.667444Z","signature_b64":"bkxgdpCQLqI3OSUPokNji/C/PqnraI+eyoz2FyQAfYJ3/Vlyw0pKY+Qa+FOAeVngYLW+tTyM3Ai0i84xv7vUBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"24d9a91e60e6d63fd3909f7ee17c1335484acb9f4b5e058a146569e7f67bd5b6","last_reissued_at":"2026-05-18T01:10:11.666986Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T01:10:11.666986Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Top-down Neural Attention by Excitation Backprop","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Jianming Zhang, Jonathan Brandt, Stan Sclaroff, Xiaohui Shen, Zhe Lin","submitted_at":"2016-08-01T17:49:57Z","abstract_excerpt":"We aim to model the top-down attention of a Convolutional Neural Network (CNN) classifier for generating task-specific attention maps. Inspired by a top-down human visual attention model, we propose a new backpropagation scheme, called Excitation Backprop, to pass along top-down signals downwards in the network hierarchy via a probabilistic Winner-Take-All process. Furthermore, we introduce the concept of contrastive attention to make the top-down attention maps more discriminative. In experiments, we demonstrate the accuracy and generalizability of our method in weakly supervised localization"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1608.00507","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":""},"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":"1608.00507","created_at":"2026-05-18T01:10:11.667058+00:00"},{"alias_kind":"arxiv_version","alias_value":"1608.00507v1","created_at":"2026-05-18T01:10:11.667058+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1608.00507","created_at":"2026-05-18T01:10:11.667058+00:00"},{"alias_kind":"pith_short_12","alias_value":"ETM2SHTA43LD","created_at":"2026-05-18T12:30:15.759754+00:00"},{"alias_kind":"pith_short_16","alias_value":"ETM2SHTA43LD7U4Q","created_at":"2026-05-18T12:30:15.759754+00:00"},{"alias_kind":"pith_short_8","alias_value":"ETM2SHTA","created_at":"2026-05-18T12:30:15.759754+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.11304","citing_title":"CheXTemporal: A Dataset for Temporally-Grounded Reasoning in Chest Radiography","ref_index":28,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ETM2SHTA43LD7U4QT57OC7ATGV","json":"https://pith.science/pith/ETM2SHTA43LD7U4QT57OC7ATGV.json","graph_json":"https://pith.science/api/pith-number/ETM2SHTA43LD7U4QT57OC7ATGV/graph.json","events_json":"https://pith.science/api/pith-number/ETM2SHTA43LD7U4QT57OC7ATGV/events.json","paper":"https://pith.science/paper/ETM2SHTA"},"agent_actions":{"view_html":"https://pith.science/pith/ETM2SHTA43LD7U4QT57OC7ATGV","download_json":"https://pith.science/pith/ETM2SHTA43LD7U4QT57OC7ATGV.json","view_paper":"https://pith.science/paper/ETM2SHTA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1608.00507&json=true","fetch_graph":"https://pith.science/api/pith-number/ETM2SHTA43LD7U4QT57OC7ATGV/graph.json","fetch_events":"https://pith.science/api/pith-number/ETM2SHTA43LD7U4QT57OC7ATGV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ETM2SHTA43LD7U4QT57OC7ATGV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ETM2SHTA43LD7U4QT57OC7ATGV/action/storage_attestation","attest_author":"https://pith.science/pith/ETM2SHTA43LD7U4QT57OC7ATGV/action/author_attestation","sign_citation":"https://pith.science/pith/ETM2SHTA43LD7U4QT57OC7ATGV/action/citation_signature","submit_replication":"https://pith.science/pith/ETM2SHTA43LD7U4QT57OC7ATGV/action/replication_record"}},"created_at":"2026-05-18T01:10:11.667058+00:00","updated_at":"2026-05-18T01:10:11.667058+00:00"}