{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:UNN3I3MAEON6YPJ5NTRFCSKNFL","short_pith_number":"pith:UNN3I3MA","schema_version":"1.0","canonical_sha256":"a35bb46d80239bec3d3d6ce251494d2ace50c6dfd51b8c05c0273f192777252f","source":{"kind":"arxiv","id":"2103.12115","version":2},"attestation_state":"computed","paper":{"title":"End-to-End Trainable Multi-Instance Pose Estimation with Transformers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Alexander Mathis, Lucas Stoffl, Maxime Vidal","submitted_at":"2021-03-22T18:19:22Z","abstract_excerpt":"We propose an end-to-end trainable approach for multi-instance pose estimation, called POET (POse Estimation Transformer). Combining a convolutional neural network with a transformer encoder-decoder architecture, we formulate multiinstance pose estimation from images as a direct set prediction problem. Our model is able to directly regress the pose of all individuals, utilizing a bipartite matching scheme. POET is trained using a novel set-based global loss that consists of a keypoint loss, a visibility loss and a class loss. POET reasons about the relations between multiple detected individua"},"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":"2103.12115","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2021-03-22T18:19:22Z","cross_cats_sorted":[],"title_canon_sha256":"404477bcc13a5648a521d14a8531dc548fda37b9371cd69a2d83699785aeca38","abstract_canon_sha256":"9b928f2a3115df94ea48fe90a494178344a20d850eea56fa7a9a34c5c280db0b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:42:21.209076Z","signature_b64":"Zb0fiMBXSCRev0fz8wQcm1Mgdb1h5qgRk9JSrNBJYPaM6HwzOb7EtgobaqMoY2o2Ijq3linInjlE4Ws6EY8IDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a35bb46d80239bec3d3d6ce251494d2ace50c6dfd51b8c05c0273f192777252f","last_reissued_at":"2026-07-05T03:42:21.208647Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:42:21.208647Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"End-to-End Trainable Multi-Instance Pose Estimation with Transformers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Alexander Mathis, Lucas Stoffl, Maxime Vidal","submitted_at":"2021-03-22T18:19:22Z","abstract_excerpt":"We propose an end-to-end trainable approach for multi-instance pose estimation, called POET (POse Estimation Transformer). Combining a convolutional neural network with a transformer encoder-decoder architecture, we formulate multiinstance pose estimation from images as a direct set prediction problem. Our model is able to directly regress the pose of all individuals, utilizing a bipartite matching scheme. POET is trained using a novel set-based global loss that consists of a keypoint loss, a visibility loss and a class loss. POET reasons about the relations between multiple detected individua"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2103.12115","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/2103.12115/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":"2103.12115","created_at":"2026-07-05T03:42:21.208704+00:00"},{"alias_kind":"arxiv_version","alias_value":"2103.12115v2","created_at":"2026-07-05T03:42:21.208704+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2103.12115","created_at":"2026-07-05T03:42:21.208704+00:00"},{"alias_kind":"pith_short_12","alias_value":"UNN3I3MAEON6","created_at":"2026-07-05T03:42:21.208704+00:00"},{"alias_kind":"pith_short_16","alias_value":"UNN3I3MAEON6YPJ5","created_at":"2026-07-05T03:42:21.208704+00:00"},{"alias_kind":"pith_short_8","alias_value":"UNN3I3MA","created_at":"2026-07-05T03:42:21.208704+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2412.02254","citing_title":"ProbPose: A Probabilistic Approach to 2D Human Pose Estimation","ref_index":22,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/UNN3I3MAEON6YPJ5NTRFCSKNFL","json":"https://pith.science/pith/UNN3I3MAEON6YPJ5NTRFCSKNFL.json","graph_json":"https://pith.science/api/pith-number/UNN3I3MAEON6YPJ5NTRFCSKNFL/graph.json","events_json":"https://pith.science/api/pith-number/UNN3I3MAEON6YPJ5NTRFCSKNFL/events.json","paper":"https://pith.science/paper/UNN3I3MA"},"agent_actions":{"view_html":"https://pith.science/pith/UNN3I3MAEON6YPJ5NTRFCSKNFL","download_json":"https://pith.science/pith/UNN3I3MAEON6YPJ5NTRFCSKNFL.json","view_paper":"https://pith.science/paper/UNN3I3MA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2103.12115&json=true","fetch_graph":"https://pith.science/api/pith-number/UNN3I3MAEON6YPJ5NTRFCSKNFL/graph.json","fetch_events":"https://pith.science/api/pith-number/UNN3I3MAEON6YPJ5NTRFCSKNFL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UNN3I3MAEON6YPJ5NTRFCSKNFL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UNN3I3MAEON6YPJ5NTRFCSKNFL/action/storage_attestation","attest_author":"https://pith.science/pith/UNN3I3MAEON6YPJ5NTRFCSKNFL/action/author_attestation","sign_citation":"https://pith.science/pith/UNN3I3MAEON6YPJ5NTRFCSKNFL/action/citation_signature","submit_replication":"https://pith.science/pith/UNN3I3MAEON6YPJ5NTRFCSKNFL/action/replication_record"}},"created_at":"2026-07-05T03:42:21.208704+00:00","updated_at":"2026-07-05T03:42:21.208704+00:00"}