{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:IHF23QEPFDL7K6AGWGZ3NLDKYN","short_pith_number":"pith:IHF23QEP","schema_version":"1.0","canonical_sha256":"41cbadc08f28d7f57806b1b3b6ac6ac37b724496f87bb1ea26d6e4ba54cc4a72","source":{"kind":"arxiv","id":"2503.13542","version":1},"attestation_state":"computed","paper":{"title":"HAR-DoReMi: Optimizing Data Mixture for Self-Supervised Human Activity Recognition Across Heterogeneous IMU Datasets","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Kevin I-Kai Wang, Liming Chen, Lulu Ban, Mingxing Nie, Qi Qiu, Shuangjian Li, Tao Zhu, Wenyong Han, Xiangqing Lu, Yaping Wan","submitted_at":"2025-03-16T04:31:58Z","abstract_excerpt":"Cross-dataset Human Activity Recognition (HAR) suffers from limited model generalization, hindering its practical deployment. To address this critical challenge, inspired by the success of DoReMi in Large Language Models (LLMs), we introduce a data mixture optimization strategy for pre-training HAR models, aiming to improve the recognition performance across heterogeneous datasets. However, directly applying DoReMi to the HAR field encounters new challenges due to the continuous, multi-channel and intrinsic heterogeneous characteristics of IMU sensor data. To overcome these limitations, we pro"},"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":"2503.13542","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-03-16T04:31:58Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"f2875cb9d30b9c5af82aafe164bb80c5242f9064a6a06fb3966f2f298871c0df","abstract_canon_sha256":"17af1b99ef15ee142c2aee21a1b5b4dac35e3ea86b47575bf0082b19f3450b26"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:33:12.504936Z","signature_b64":"tvcZCtxPfokmujfxPHiOZswkyyYl/lz5Y29t5ZHL7eS2HBxJm1p3J44TS2w6ofNniTPEi5hDFjVqPA/zFnEMBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"41cbadc08f28d7f57806b1b3b6ac6ac37b724496f87bb1ea26d6e4ba54cc4a72","last_reissued_at":"2026-07-05T10:33:12.503987Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:33:12.503987Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"HAR-DoReMi: Optimizing Data Mixture for Self-Supervised Human Activity Recognition Across Heterogeneous IMU Datasets","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Kevin I-Kai Wang, Liming Chen, Lulu Ban, Mingxing Nie, Qi Qiu, Shuangjian Li, Tao Zhu, Wenyong Han, Xiangqing Lu, Yaping Wan","submitted_at":"2025-03-16T04:31:58Z","abstract_excerpt":"Cross-dataset Human Activity Recognition (HAR) suffers from limited model generalization, hindering its practical deployment. To address this critical challenge, inspired by the success of DoReMi in Large Language Models (LLMs), we introduce a data mixture optimization strategy for pre-training HAR models, aiming to improve the recognition performance across heterogeneous datasets. However, directly applying DoReMi to the HAR field encounters new challenges due to the continuous, multi-channel and intrinsic heterogeneous characteristics of IMU sensor data. To overcome these limitations, we pro"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.13542","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/2503.13542/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":"2503.13542","created_at":"2026-07-05T10:33:12.504113+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.13542v1","created_at":"2026-07-05T10:33:12.504113+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.13542","created_at":"2026-07-05T10:33:12.504113+00:00"},{"alias_kind":"pith_short_12","alias_value":"IHF23QEPFDL7","created_at":"2026-07-05T10:33:12.504113+00:00"},{"alias_kind":"pith_short_16","alias_value":"IHF23QEPFDL7K6AG","created_at":"2026-07-05T10:33:12.504113+00:00"},{"alias_kind":"pith_short_8","alias_value":"IHF23QEP","created_at":"2026-07-05T10:33:12.504113+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.02711","citing_title":"Foundation Models Defining A New Era In Sensor-based Human Activity Recognition: A Survey And Outlook","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2604.21369","citing_title":"Channel-Free Human Activity Recognition via Inductive-Bias-Aware Fusion Design for Heterogeneous IoT Sensor Environments","ref_index":18,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/IHF23QEPFDL7K6AGWGZ3NLDKYN","json":"https://pith.science/pith/IHF23QEPFDL7K6AGWGZ3NLDKYN.json","graph_json":"https://pith.science/api/pith-number/IHF23QEPFDL7K6AGWGZ3NLDKYN/graph.json","events_json":"https://pith.science/api/pith-number/IHF23QEPFDL7K6AGWGZ3NLDKYN/events.json","paper":"https://pith.science/paper/IHF23QEP"},"agent_actions":{"view_html":"https://pith.science/pith/IHF23QEPFDL7K6AGWGZ3NLDKYN","download_json":"https://pith.science/pith/IHF23QEPFDL7K6AGWGZ3NLDKYN.json","view_paper":"https://pith.science/paper/IHF23QEP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.13542&json=true","fetch_graph":"https://pith.science/api/pith-number/IHF23QEPFDL7K6AGWGZ3NLDKYN/graph.json","fetch_events":"https://pith.science/api/pith-number/IHF23QEPFDL7K6AGWGZ3NLDKYN/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IHF23QEPFDL7K6AGWGZ3NLDKYN/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IHF23QEPFDL7K6AGWGZ3NLDKYN/action/storage_attestation","attest_author":"https://pith.science/pith/IHF23QEPFDL7K6AGWGZ3NLDKYN/action/author_attestation","sign_citation":"https://pith.science/pith/IHF23QEPFDL7K6AGWGZ3NLDKYN/action/citation_signature","submit_replication":"https://pith.science/pith/IHF23QEPFDL7K6AGWGZ3NLDKYN/action/replication_record"}},"created_at":"2026-07-05T10:33:12.504113+00:00","updated_at":"2026-07-05T10:33:12.504113+00:00"}