{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:LNV5F5DANXWAYSN25XW4MAGDVZ","short_pith_number":"pith:LNV5F5DA","canonical_record":{"source":{"id":"2605.13604","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-05-13T14:39:24Z","cross_cats_sorted":[],"title_canon_sha256":"13a4a3dbc6627a86e2f0cfe61121c1938268ef673b5c7b6320f01d0695c7708c","abstract_canon_sha256":"9a54a6abbe50aa07d7731aa3d34baf44e714d139388e186bec5500732c3ac533"},"schema_version":"1.0"},"canonical_sha256":"5b6bd2f4606dec0c49baededc600c3ae5bdc24a7cc1b522340ea1c10710851e3","source":{"kind":"arxiv","id":"2605.13604","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.13604","created_at":"2026-05-18T02:44:22Z"},{"alias_kind":"arxiv_version","alias_value":"2605.13604v1","created_at":"2026-05-18T02:44:22Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.13604","created_at":"2026-05-18T02:44:22Z"},{"alias_kind":"pith_short_12","alias_value":"LNV5F5DANXWA","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"LNV5F5DANXWAYSN2","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"LNV5F5DA","created_at":"2026-05-18T12:33:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:LNV5F5DANXWAYSN25XW4MAGDVZ","target":"record","payload":{"canonical_record":{"source":{"id":"2605.13604","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-05-13T14:39:24Z","cross_cats_sorted":[],"title_canon_sha256":"13a4a3dbc6627a86e2f0cfe61121c1938268ef673b5c7b6320f01d0695c7708c","abstract_canon_sha256":"9a54a6abbe50aa07d7731aa3d34baf44e714d139388e186bec5500732c3ac533"},"schema_version":"1.0"},"canonical_sha256":"5b6bd2f4606dec0c49baededc600c3ae5bdc24a7cc1b522340ea1c10710851e3","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T02:44:22.887419Z","signature_b64":"fY8IAmPSLZllBct2dJBiyM6wQJfLGJ8xDXuvzzD2dTol0lW2e+50/JXhaE7klUOxG3jBmgMw1QXtPY1LnH2RCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5b6bd2f4606dec0c49baededc600c3ae5bdc24a7cc1b522340ea1c10710851e3","last_reissued_at":"2026-05-18T02:44:22.887004Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T02:44:22.887004Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.13604","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T02:44:22Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2lSZDfjvfqUbhPHv3vOebD9i0LlNW9OodoOxW3EVcvkPw1yLq7peBltqpA/VotcwKPUOLYogovoKe+b7xev/Dg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-26T10:35:29.595992Z"},"content_sha256":"f23578916244e8a6ed37ead3baabb05f05b35b5489dcaca4afbf8717ca6f57de","schema_version":"1.0","event_id":"sha256:f23578916244e8a6ed37ead3baabb05f05b35b5489dcaca4afbf8717ca6f57de"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:LNV5F5DANXWAYSN25XW4MAGDVZ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Rethinking Graph Convolution for 2D-to-3D Hand Pose Lifting","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Adaptive attention outperforms fixed graph convolution for lifting 2D hand poses to 3D.","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chanyoung Kim, Donghyun Kim, Dong-Hyun Sim, Seong Jae Hwang, Youngjoong Kwon","submitted_at":"2026-05-13T14:39:24Z","abstract_excerpt":"Graph convolutional networks (GCNs) are widely used for 3D hand pose estimation, where the hand skeleton is encoded as a fixed adjacency graph. We revisit whether this is the most effective way to incorporate hand topology in 2D-to-3D lifting. In this paper, we perform controlled, parameter-matched ablations on the FPHA benchmark and show that standard multi-head self-attention consistently outperforms GCN baselines. Even when the GCN is strengthened with multi-hop adjacency and matched parameter count, self-attention reduces MPJPE from 12.36 mm to 10.09 mm. A skeleton-constrained graph attent"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"These results suggest that, for hand pose lifting, adaptive spatial attention is a more effective inductive bias than fixed graph convolution.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The assumption that parameter-matched ablations on the FPHA benchmark alone establish the general superiority of attention over GCNs across hand pose lifting tasks and datasets.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Self-attention with input-dependent aggregation and soft graph-distance priors outperforms fixed graph convolutions for 2D-to-3D hand pose estimation on FPHA.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Adaptive attention outperforms fixed graph convolution for lifting 2D hand poses to 3D.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"ecc16362c7c194e2361129087e9ad0e56b753880d1b6ad97573bcf9d35c7776e"},"source":{"id":"2605.13604","kind":"arxiv","version":1},"verdict":{"id":"e6ccb514-8b9b-4bf7-8902-ccad8de72b85","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-14T19:18:04.208070Z","strongest_claim":"These results suggest that, for hand pose lifting, adaptive spatial attention is a more effective inductive bias than fixed graph convolution.","one_line_summary":"Self-attention with input-dependent aggregation and soft graph-distance priors outperforms fixed graph convolutions for 2D-to-3D hand pose estimation on FPHA.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The assumption that parameter-matched ablations on the FPHA benchmark alone establish the general superiority of attention over GCNs across hand pose lifting tasks and datasets.","pith_extraction_headline":"Adaptive attention outperforms fixed graph convolution for lifting 2D hand poses to 3D."},"references":{"count":14,"sample":[{"doi":"","year":2018,"title":"First-person hand action bench- mark with rgb-d videos and 3d hand pose annotations","work_id":"4fcf3365-5d83-47df-8634-ef8ae945f996","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2017,"title":"Kipf and Max Welling","work_id":"5ad16f04-d1f4-40f0-a90d-8f54486aa7a3","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"Gtignet: Global topology interaction graphormer network for 3d hand pose estimation.Neural Networks, 185:107221, 2025","work_id":"24ac1926-d62f-4442-a761-10d46a656967","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2019,"title":"Decoupled weight de- cay regularization","work_id":"f3b247f3-b210-4494-b445-0338427b4292","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Assemblyhands: Towards egocen- tric activity understanding via 3d hand pose estimation","work_id":"19eec250-2859-4986-a38c-c11fb2b1db68","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":14,"snapshot_sha256":"8303fbb739e907182f01eefd98113ac9462e0709e5bb6f519f237853f30f5d19","internal_anchors":0},"formal_canon":{"evidence_count":2,"snapshot_sha256":"3832dece7c7ef158b4995356e1a24dd2d26d1bbde4710d1090e8a51717416f87"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"e6ccb514-8b9b-4bf7-8902-ccad8de72b85"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T02:44:22Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"9ZgrRQ4SL88xlQJcs8xXjws81Tc8VKVBrzen25myz8xOEwfCXhtP5bM5CVXDKtPwrQOFhylIuHA3cldQvgWQAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-26T10:35:29.596650Z"},"content_sha256":"cca7913a3288d7177ab4956cb5910cdfd2f2748a3fcd873524320c2db3023e4e","schema_version":"1.0","event_id":"sha256:cca7913a3288d7177ab4956cb5910cdfd2f2748a3fcd873524320c2db3023e4e"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/LNV5F5DANXWAYSN25XW4MAGDVZ/bundle.json","state_url":"https://pith.science/pith/LNV5F5DANXWAYSN25XW4MAGDVZ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/LNV5F5DANXWAYSN25XW4MAGDVZ/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-05-26T10:35:29Z","links":{"resolver":"https://pith.science/pith/LNV5F5DANXWAYSN25XW4MAGDVZ","bundle":"https://pith.science/pith/LNV5F5DANXWAYSN25XW4MAGDVZ/bundle.json","state":"https://pith.science/pith/LNV5F5DANXWAYSN25XW4MAGDVZ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/LNV5F5DANXWAYSN25XW4MAGDVZ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:LNV5F5DANXWAYSN25XW4MAGDVZ","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"9a54a6abbe50aa07d7731aa3d34baf44e714d139388e186bec5500732c3ac533","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-05-13T14:39:24Z","title_canon_sha256":"13a4a3dbc6627a86e2f0cfe61121c1938268ef673b5c7b6320f01d0695c7708c"},"schema_version":"1.0","source":{"id":"2605.13604","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.13604","created_at":"2026-05-18T02:44:22Z"},{"alias_kind":"arxiv_version","alias_value":"2605.13604v1","created_at":"2026-05-18T02:44:22Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.13604","created_at":"2026-05-18T02:44:22Z"},{"alias_kind":"pith_short_12","alias_value":"LNV5F5DANXWA","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"LNV5F5DANXWAYSN2","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"LNV5F5DA","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:cca7913a3288d7177ab4956cb5910cdfd2f2748a3fcd873524320c2db3023e4e","target":"graph","created_at":"2026-05-18T02:44:22Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"These results suggest that, for hand pose lifting, adaptive spatial attention is a more effective inductive bias than fixed graph convolution."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"The assumption that parameter-matched ablations on the FPHA benchmark alone establish the general superiority of attention over GCNs across hand pose lifting tasks and datasets."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"Self-attention with input-dependent aggregation and soft graph-distance priors outperforms fixed graph convolutions for 2D-to-3D hand pose estimation on FPHA."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Adaptive attention outperforms fixed graph convolution for lifting 2D hand poses to 3D."}],"snapshot_sha256":"ecc16362c7c194e2361129087e9ad0e56b753880d1b6ad97573bcf9d35c7776e"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"3832dece7c7ef158b4995356e1a24dd2d26d1bbde4710d1090e8a51717416f87"},"paper":{"abstract_excerpt":"Graph convolutional networks (GCNs) are widely used for 3D hand pose estimation, where the hand skeleton is encoded as a fixed adjacency graph. We revisit whether this is the most effective way to incorporate hand topology in 2D-to-3D lifting. In this paper, we perform controlled, parameter-matched ablations on the FPHA benchmark and show that standard multi-head self-attention consistently outperforms GCN baselines. Even when the GCN is strengthened with multi-hop adjacency and matched parameter count, self-attention reduces MPJPE from 12.36 mm to 10.09 mm. A skeleton-constrained graph attent","authors_text":"Chanyoung Kim, Donghyun Kim, Dong-Hyun Sim, Seong Jae Hwang, Youngjoong Kwon","cross_cats":[],"headline":"Adaptive attention outperforms fixed graph convolution for lifting 2D hand poses to 3D.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-05-13T14:39:24Z","title":"Rethinking Graph Convolution for 2D-to-3D Hand Pose Lifting"},"references":{"count":14,"internal_anchors":0,"resolved_work":14,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"First-person hand action bench- mark with rgb-d videos and 3d hand pose annotations","work_id":"4fcf3365-5d83-47df-8634-ef8ae945f996","year":2018},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"Kipf and Max Welling","work_id":"5ad16f04-d1f4-40f0-a90d-8f54486aa7a3","year":2017},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"Gtignet: Global topology interaction graphormer network for 3d hand pose estimation.Neural Networks, 185:107221, 2025","work_id":"24ac1926-d62f-4442-a761-10d46a656967","year":2025},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"Decoupled weight de- cay regularization","work_id":"f3b247f3-b210-4494-b445-0338427b4292","year":2019},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"Assemblyhands: Towards egocen- tric activity understanding via 3d hand pose estimation","work_id":"19eec250-2859-4986-a38c-c11fb2b1db68","year":null}],"snapshot_sha256":"8303fbb739e907182f01eefd98113ac9462e0709e5bb6f519f237853f30f5d19"},"source":{"id":"2605.13604","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-14T19:18:04.208070Z","id":"e6ccb514-8b9b-4bf7-8902-ccad8de72b85","model_set":{"reader":"grok-4.3"},"one_line_summary":"Self-attention with input-dependent aggregation and soft graph-distance priors outperforms fixed graph convolutions for 2D-to-3D hand pose estimation on FPHA.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Adaptive attention outperforms fixed graph convolution for lifting 2D hand poses to 3D.","strongest_claim":"These results suggest that, for hand pose lifting, adaptive spatial attention is a more effective inductive bias than fixed graph convolution.","weakest_assumption":"The assumption that parameter-matched ablations on the FPHA benchmark alone establish the general superiority of attention over GCNs across hand pose lifting tasks and datasets."}},"verdict_id":"e6ccb514-8b9b-4bf7-8902-ccad8de72b85"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:f23578916244e8a6ed37ead3baabb05f05b35b5489dcaca4afbf8717ca6f57de","target":"record","created_at":"2026-05-18T02:44:22Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"9a54a6abbe50aa07d7731aa3d34baf44e714d139388e186bec5500732c3ac533","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-05-13T14:39:24Z","title_canon_sha256":"13a4a3dbc6627a86e2f0cfe61121c1938268ef673b5c7b6320f01d0695c7708c"},"schema_version":"1.0","source":{"id":"2605.13604","kind":"arxiv","version":1}},"canonical_sha256":"5b6bd2f4606dec0c49baededc600c3ae5bdc24a7cc1b522340ea1c10710851e3","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"5b6bd2f4606dec0c49baededc600c3ae5bdc24a7cc1b522340ea1c10710851e3","first_computed_at":"2026-05-18T02:44:22.887004Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T02:44:22.887004Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"fY8IAmPSLZllBct2dJBiyM6wQJfLGJ8xDXuvzzD2dTol0lW2e+50/JXhaE7klUOxG3jBmgMw1QXtPY1LnH2RCg==","signature_status":"signed_v1","signed_at":"2026-05-18T02:44:22.887419Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.13604","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f23578916244e8a6ed37ead3baabb05f05b35b5489dcaca4afbf8717ca6f57de","sha256:cca7913a3288d7177ab4956cb5910cdfd2f2748a3fcd873524320c2db3023e4e"],"state_sha256":"828d7935dfd4a98d0b86fa5f5852fdb72bea5ce1383ee522ee8754e11a343e9c"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"sKam+xH5YCGt+glrpnRUK8cykTcenCWT6bxna5ztK1zvtyh+bZ3HVpejoIeMBkaX/Wjb14bQK4JaLX11tUkdAg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-26T10:35:29.600664Z","bundle_sha256":"498da87154f68282a77791dadd6ec63167e508a9db1a2b17e101734dab8a0934"}}