{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:QMTMYA7HX4DFIPUZYKHH65PBNG","short_pith_number":"pith:QMTMYA7H","schema_version":"1.0","canonical_sha256":"8326cc03e7bf06543e99c28e7f75e169a1d8e336eac4c29b28c0649c58a2d40f","source":{"kind":"arxiv","id":"2112.02439","version":1},"attestation_state":"computed","paper":{"title":"Deep Learning on Mobile Devices Through Neural Processing Units and Edge Computing","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.NI","authors_text":"Guohong Cao, Tianxiang Tan","submitted_at":"2021-12-04T22:04:21Z","abstract_excerpt":"Deep Neural Network (DNN) is becoming adopted for video analytics on mobile devices. To reduce the delay of running DNNs, many mobile devices are equipped with Neural Processing Units (NPU). However, due to the resource limitations of NPU, these DNNs have to be compressed to increase the processing speed at the cost of accuracy. To address the low accuracy problem, we propose a Confidence Based Offloading (CBO) framework for deep learning video analytics. The major challenge is to determine when to return the NPU classification result based on the confidence level of running the DNN, and when "},"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":"2112.02439","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.NI","submitted_at":"2021-12-04T22:04:21Z","cross_cats_sorted":[],"title_canon_sha256":"e7211cb83fad7d4e858b11670bec16d1b91eeee6b557d94a32317f5dc6648300","abstract_canon_sha256":"38c32b96f3662074660bb0b19d150e71a4e113a276314ade25f5064faa80ef8a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:37:52.815474Z","signature_b64":"ZNzB4EmnrTN77ifDx/U8CirA3GzHAYG0G9ep3sKlYr+45e4c7wt8Om8d2rkqHkA8u65WgSIaS382SKIybvI8DA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8326cc03e7bf06543e99c28e7f75e169a1d8e336eac4c29b28c0649c58a2d40f","last_reissued_at":"2026-07-05T03:37:52.815079Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:37:52.815079Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Deep Learning on Mobile Devices Through Neural Processing Units and Edge Computing","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.NI","authors_text":"Guohong Cao, Tianxiang Tan","submitted_at":"2021-12-04T22:04:21Z","abstract_excerpt":"Deep Neural Network (DNN) is becoming adopted for video analytics on mobile devices. To reduce the delay of running DNNs, many mobile devices are equipped with Neural Processing Units (NPU). However, due to the resource limitations of NPU, these DNNs have to be compressed to increase the processing speed at the cost of accuracy. To address the low accuracy problem, we propose a Confidence Based Offloading (CBO) framework for deep learning video analytics. The major challenge is to determine when to return the NPU classification result based on the confidence level of running the DNN, and when "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2112.02439","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/2112.02439/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":"2112.02439","created_at":"2026-07-05T03:37:52.815134+00:00"},{"alias_kind":"arxiv_version","alias_value":"2112.02439v1","created_at":"2026-07-05T03:37:52.815134+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2112.02439","created_at":"2026-07-05T03:37:52.815134+00:00"},{"alias_kind":"pith_short_12","alias_value":"QMTMYA7HX4DF","created_at":"2026-07-05T03:37:52.815134+00:00"},{"alias_kind":"pith_short_16","alias_value":"QMTMYA7HX4DFIPUZ","created_at":"2026-07-05T03:37:52.815134+00:00"},{"alias_kind":"pith_short_8","alias_value":"QMTMYA7H","created_at":"2026-07-05T03:37:52.815134+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2505.11417","citing_title":"EdgeWisePersona: A Dataset for On-Device User Profiling from Natural Language Interactions","ref_index":31,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QMTMYA7HX4DFIPUZYKHH65PBNG","json":"https://pith.science/pith/QMTMYA7HX4DFIPUZYKHH65PBNG.json","graph_json":"https://pith.science/api/pith-number/QMTMYA7HX4DFIPUZYKHH65PBNG/graph.json","events_json":"https://pith.science/api/pith-number/QMTMYA7HX4DFIPUZYKHH65PBNG/events.json","paper":"https://pith.science/paper/QMTMYA7H"},"agent_actions":{"view_html":"https://pith.science/pith/QMTMYA7HX4DFIPUZYKHH65PBNG","download_json":"https://pith.science/pith/QMTMYA7HX4DFIPUZYKHH65PBNG.json","view_paper":"https://pith.science/paper/QMTMYA7H","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2112.02439&json=true","fetch_graph":"https://pith.science/api/pith-number/QMTMYA7HX4DFIPUZYKHH65PBNG/graph.json","fetch_events":"https://pith.science/api/pith-number/QMTMYA7HX4DFIPUZYKHH65PBNG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QMTMYA7HX4DFIPUZYKHH65PBNG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QMTMYA7HX4DFIPUZYKHH65PBNG/action/storage_attestation","attest_author":"https://pith.science/pith/QMTMYA7HX4DFIPUZYKHH65PBNG/action/author_attestation","sign_citation":"https://pith.science/pith/QMTMYA7HX4DFIPUZYKHH65PBNG/action/citation_signature","submit_replication":"https://pith.science/pith/QMTMYA7HX4DFIPUZYKHH65PBNG/action/replication_record"}},"created_at":"2026-07-05T03:37:52.815134+00:00","updated_at":"2026-07-05T03:37:52.815134+00:00"}