{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:3LWEE4NEODIPOQCPA555C3HA2T","short_pith_number":"pith:3LWEE4NE","schema_version":"1.0","canonical_sha256":"daec4271a470d0f7404f077bd16ce0d4f8bc13ac6480137ec9d8c78d602699b5","source":{"kind":"arxiv","id":"2210.11179","version":1},"attestation_state":"computed","paper":{"title":"Multi-hypothesis 3D human pose estimation metrics favor miscalibrated distributions","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Fabian H. Sinz, Mohammad Bashiri, Pawe{\\l} A. Pierzchlewicz, R. James Cotton","submitted_at":"2022-10-20T11:47:07Z","abstract_excerpt":"Due to depth ambiguities and occlusions, lifting 2D poses to 3D is a highly ill-posed problem. Well-calibrated distributions of possible poses can make these ambiguities explicit and preserve the resulting uncertainty for downstream tasks. This study shows that previous attempts, which account for these ambiguities via multiple hypotheses generation, produce miscalibrated distributions. We identify that miscalibration can be attributed to the use of sample-based metrics such as minMPJPE. In a series of simulations, we show that minimizing minMPJPE, as commonly done, should converge to the corr"},"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":"2210.11179","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2022-10-20T11:47:07Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"8045bdd14e05520605996d0012f2fdef61b1bbc03faaaa7f83f954599a70dee9","abstract_canon_sha256":"9617c91241d910ed0a1e5201b2ed10f7562554fd224af8d8e554afd7ff23c413"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:08:45.375855Z","signature_b64":"zpNisiKRgY835VAqR0fMUWfpRkuffFtilr5KQ/aAiKl2K3VmLDgd69lmehMuHMOM6UY8TxUKhMu8CqaUB2r8DA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"daec4271a470d0f7404f077bd16ce0d4f8bc13ac6480137ec9d8c78d602699b5","last_reissued_at":"2026-07-05T05:08:45.375469Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:08:45.375469Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Multi-hypothesis 3D human pose estimation metrics favor miscalibrated distributions","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Fabian H. Sinz, Mohammad Bashiri, Pawe{\\l} A. Pierzchlewicz, R. James Cotton","submitted_at":"2022-10-20T11:47:07Z","abstract_excerpt":"Due to depth ambiguities and occlusions, lifting 2D poses to 3D is a highly ill-posed problem. Well-calibrated distributions of possible poses can make these ambiguities explicit and preserve the resulting uncertainty for downstream tasks. This study shows that previous attempts, which account for these ambiguities via multiple hypotheses generation, produce miscalibrated distributions. We identify that miscalibration can be attributed to the use of sample-based metrics such as minMPJPE. In a series of simulations, we show that minimizing minMPJPE, as commonly done, should converge to the corr"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2210.11179","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/2210.11179/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":"2210.11179","created_at":"2026-07-05T05:08:45.375524+00:00"},{"alias_kind":"arxiv_version","alias_value":"2210.11179v1","created_at":"2026-07-05T05:08:45.375524+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2210.11179","created_at":"2026-07-05T05:08:45.375524+00:00"},{"alias_kind":"pith_short_12","alias_value":"3LWEE4NEODIP","created_at":"2026-07-05T05:08:45.375524+00:00"},{"alias_kind":"pith_short_16","alias_value":"3LWEE4NEODIPOQCP","created_at":"2026-07-05T05:08:45.375524+00:00"},{"alias_kind":"pith_short_8","alias_value":"3LWEE4NE","created_at":"2026-07-05T05:08:45.375524+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2608.09579","citing_title":"You Only Flow Once: Calibrated and Real-Time Radar Pose Estimation with Multi-Hypothesis Normalizing Flows","ref_index":36,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3LWEE4NEODIPOQCPA555C3HA2T","json":"https://pith.science/pith/3LWEE4NEODIPOQCPA555C3HA2T.json","graph_json":"https://pith.science/api/pith-number/3LWEE4NEODIPOQCPA555C3HA2T/graph.json","events_json":"https://pith.science/api/pith-number/3LWEE4NEODIPOQCPA555C3HA2T/events.json","paper":"https://pith.science/paper/3LWEE4NE"},"agent_actions":{"view_html":"https://pith.science/pith/3LWEE4NEODIPOQCPA555C3HA2T","download_json":"https://pith.science/pith/3LWEE4NEODIPOQCPA555C3HA2T.json","view_paper":"https://pith.science/paper/3LWEE4NE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2210.11179&json=true","fetch_graph":"https://pith.science/api/pith-number/3LWEE4NEODIPOQCPA555C3HA2T/graph.json","fetch_events":"https://pith.science/api/pith-number/3LWEE4NEODIPOQCPA555C3HA2T/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3LWEE4NEODIPOQCPA555C3HA2T/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3LWEE4NEODIPOQCPA555C3HA2T/action/storage_attestation","attest_author":"https://pith.science/pith/3LWEE4NEODIPOQCPA555C3HA2T/action/author_attestation","sign_citation":"https://pith.science/pith/3LWEE4NEODIPOQCPA555C3HA2T/action/citation_signature","submit_replication":"https://pith.science/pith/3LWEE4NEODIPOQCPA555C3HA2T/action/replication_record"}},"created_at":"2026-07-05T05:08:45.375524+00:00","updated_at":"2026-07-05T05:08:45.375524+00:00"}