{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:C5GU3BJRY7XMSKMQZ3G7DV7V4S","short_pith_number":"pith:C5GU3BJR","schema_version":"1.0","canonical_sha256":"174d4d8531c7eec92990cecdf1d7f5e4bb47eb29c099c79457313f36524e728e","source":{"kind":"arxiv","id":"2304.03768","version":1},"attestation_state":"computed","paper":{"title":"SparseFormer: Sparse Visual Recognition via Limited Latent Tokens","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Limin Wang, Mike Zheng Shou, Zhan Tong, Ziteng Gao","submitted_at":"2023-04-07T17:59:58Z","abstract_excerpt":"Human visual recognition is a sparse process, where only a few salient visual cues are attended to rather than traversing every detail uniformly. However, most current vision networks follow a dense paradigm, processing every single visual unit (e.g,, pixel or patch) in a uniform manner. In this paper, we challenge this dense paradigm and present a new method, coined SparseFormer, to imitate human's sparse visual recognition in an end-to-end manner. SparseFormer learns to represent images using a highly limited number of tokens (down to 49) in the latent space with sparse feature sampling proc"},"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":"2304.03768","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2023-04-07T17:59:58Z","cross_cats_sorted":[],"title_canon_sha256":"8225579b5d07c5cec0947a367c2962aff6181889067fbae1b6c086902025dbe4","abstract_canon_sha256":"734165869ff7d655d8d354c50cea698af05a958384b619e207cd200dc829b408"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:59:00.168743Z","signature_b64":"neItAyR/yPt0r0Imdjzz6/EB2uemJLotcxiM67AU9vXLeJ8UL5WC17koYb/w9H04/pQgZHqt/hvYeAd2Tla/Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"174d4d8531c7eec92990cecdf1d7f5e4bb47eb29c099c79457313f36524e728e","last_reissued_at":"2026-07-05T05:59:00.168332Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:59:00.168332Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SparseFormer: Sparse Visual Recognition via Limited Latent Tokens","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Limin Wang, Mike Zheng Shou, Zhan Tong, Ziteng Gao","submitted_at":"2023-04-07T17:59:58Z","abstract_excerpt":"Human visual recognition is a sparse process, where only a few salient visual cues are attended to rather than traversing every detail uniformly. However, most current vision networks follow a dense paradigm, processing every single visual unit (e.g,, pixel or patch) in a uniform manner. In this paper, we challenge this dense paradigm and present a new method, coined SparseFormer, to imitate human's sparse visual recognition in an end-to-end manner. SparseFormer learns to represent images using a highly limited number of tokens (down to 49) in the latent space with sparse feature sampling proc"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2304.03768","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/2304.03768/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":"2304.03768","created_at":"2026-07-05T05:59:00.168387+00:00"},{"alias_kind":"arxiv_version","alias_value":"2304.03768v1","created_at":"2026-07-05T05:59:00.168387+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2304.03768","created_at":"2026-07-05T05:59:00.168387+00:00"},{"alias_kind":"pith_short_12","alias_value":"C5GU3BJRY7XM","created_at":"2026-07-05T05:59:00.168387+00:00"},{"alias_kind":"pith_short_16","alias_value":"C5GU3BJRY7XMSKMQ","created_at":"2026-07-05T05:59:00.168387+00:00"},{"alias_kind":"pith_short_8","alias_value":"C5GU3BJR","created_at":"2026-07-05T05:59:00.168387+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2501.16811","citing_title":"Not Every Patch is Needed: Towards a More Efficient and Effective Backbone for Video-based Person Re-identification","ref_index":21,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/C5GU3BJRY7XMSKMQZ3G7DV7V4S","json":"https://pith.science/pith/C5GU3BJRY7XMSKMQZ3G7DV7V4S.json","graph_json":"https://pith.science/api/pith-number/C5GU3BJRY7XMSKMQZ3G7DV7V4S/graph.json","events_json":"https://pith.science/api/pith-number/C5GU3BJRY7XMSKMQZ3G7DV7V4S/events.json","paper":"https://pith.science/paper/C5GU3BJR"},"agent_actions":{"view_html":"https://pith.science/pith/C5GU3BJRY7XMSKMQZ3G7DV7V4S","download_json":"https://pith.science/pith/C5GU3BJRY7XMSKMQZ3G7DV7V4S.json","view_paper":"https://pith.science/paper/C5GU3BJR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2304.03768&json=true","fetch_graph":"https://pith.science/api/pith-number/C5GU3BJRY7XMSKMQZ3G7DV7V4S/graph.json","fetch_events":"https://pith.science/api/pith-number/C5GU3BJRY7XMSKMQZ3G7DV7V4S/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/C5GU3BJRY7XMSKMQZ3G7DV7V4S/action/timestamp_anchor","attest_storage":"https://pith.science/pith/C5GU3BJRY7XMSKMQZ3G7DV7V4S/action/storage_attestation","attest_author":"https://pith.science/pith/C5GU3BJRY7XMSKMQZ3G7DV7V4S/action/author_attestation","sign_citation":"https://pith.science/pith/C5GU3BJRY7XMSKMQZ3G7DV7V4S/action/citation_signature","submit_replication":"https://pith.science/pith/C5GU3BJRY7XMSKMQZ3G7DV7V4S/action/replication_record"}},"created_at":"2026-07-05T05:59:00.168387+00:00","updated_at":"2026-07-05T05:59:00.168387+00:00"}