{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:4FRWERRYJFDTTKL6ILZ24KR4LZ","short_pith_number":"pith:4FRWERRY","schema_version":"1.0","canonical_sha256":"e163624638494739a97e42f3ae2a3c5e5bee0c6a099dde5c501a5eb1c80ffb92","source":{"kind":"arxiv","id":"2605.15423","version":1},"attestation_state":"computed","paper":{"title":"MR2-ByteTrack: CNN and Transformer-based Video Object Detection for AI-augmented Embedded Vision Sensor Nodes","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"MR2-ByteTrack enables video object detection with up to 55% energy savings on microcontroller-based vision sensors by alternating resolutions and rescoring detections.","cross_cats":["cs.AI","eess.IV"],"primary_cat":"cs.CV","authors_text":"Daniele Palossi, Francesco Conti, Luca Benini, Luca Bompani, Manuele Rusci","submitted_at":"2026-05-14T21:13:56Z","abstract_excerpt":"Modern smart vision sensors need on-device intelligence to process video streams, as cloud computing is often impractical due to bandwidth, latency, and privacy constraints. However, these sensory systems typically rely on ultra-low-power microcontrollers (MCUs) with limited memory and compute, making conventional video object detection methods, which require feature storage or multi-frame buffering, unfeasible. To address this challenge, we introduce Multi-Resolution Rescored ByteTrack (MR2-ByteTrack), a Video Object Detection (VOD) method tailored for MCU-based embedded vision nodes. MR2-Byt"},"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":true,"formal_links_present":true},"canonical_record":{"source":{"id":"2605.15423","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2026-05-14T21:13:56Z","cross_cats_sorted":["cs.AI","eess.IV"],"title_canon_sha256":"9a8b912843987b034c5ca861ddec9bbecc940db33e27dafd3079589d9008adf4","abstract_canon_sha256":"b41e1ac4e50b9c18a4b516f21ea057eae9c978954984f211bf6a769a5358c633"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:00:57.845439Z","signature_b64":"Vb/s01+iUzCAFjo6WJwENp37BMUim60Dt7clx16MTTeXmeWRRdQgl+bNV1/rBSnneIIlM6Xo3k51Z4gky1/IDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e163624638494739a97e42f3ae2a3c5e5bee0c6a099dde5c501a5eb1c80ffb92","last_reissued_at":"2026-05-20T00:00:57.844623Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:00:57.844623Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MR2-ByteTrack: CNN and Transformer-based Video Object Detection for AI-augmented Embedded Vision Sensor Nodes","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"MR2-ByteTrack enables video object detection with up to 55% energy savings on microcontroller-based vision sensors by alternating resolutions and rescoring detections.","cross_cats":["cs.AI","eess.IV"],"primary_cat":"cs.CV","authors_text":"Daniele Palossi, Francesco Conti, Luca Benini, Luca Bompani, Manuele Rusci","submitted_at":"2026-05-14T21:13:56Z","abstract_excerpt":"Modern smart vision sensors need on-device intelligence to process video streams, as cloud computing is often impractical due to bandwidth, latency, and privacy constraints. However, these sensory systems typically rely on ultra-low-power microcontrollers (MCUs) with limited memory and compute, making conventional video object detection methods, which require feature storage or multi-frame buffering, unfeasible. To address this challenge, we introduce Multi-Resolution Rescored ByteTrack (MR2-ByteTrack), a Video Object Detection (VOD) method tailored for MCU-based embedded vision nodes. MR2-Byt"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"our method yields up to 55% energy savings compared to processing only full-resolution images, enabling the first real-time Transformer-based VOD on an MCU-class embedded vision node.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The Rescore algorithm, which applies probability union rules to aggregate detection confidence scores across frames, can reliably correct misclassifications from low-resolution inferences without degrading overall performance.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"MR2-ByteTrack maintains high accuracy in video object detection on MCUs by combining multi-resolution processing, ByteTrack for frame linking, and Rescore for confidence aggregation, achieving up to 55% energy savings and real-time performance for both CNN and Transformer models.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"MR2-ByteTrack enables video object detection with up to 55% energy savings on microcontroller-based vision sensors by alternating resolutions and rescoring detections.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"2180dee21ad4a78081f79ac03e737ebfab3b0c4aa68bb52c8ee46d54dc925f64"},"source":{"id":"2605.15423","kind":"arxiv","version":1},"verdict":{"id":"d1fb2793-35f1-42dc-a929-722e7972751c","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-19T15:20:15.327680Z","strongest_claim":"our method yields up to 55% energy savings compared to processing only full-resolution images, enabling the first real-time Transformer-based VOD on an MCU-class embedded vision node.","one_line_summary":"MR2-ByteTrack maintains high accuracy in video object detection on MCUs by combining multi-resolution processing, ByteTrack for frame linking, and Rescore for confidence aggregation, achieving up to 55% energy savings and real-time performance for both CNN and Transformer models.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The Rescore algorithm, which applies probability union rules to aggregate detection confidence scores across frames, can reliably correct misclassifications from low-resolution inferences without degrading overall performance.","pith_extraction_headline":"MR2-ByteTrack enables video object detection with up to 55% energy savings on microcontroller-based vision sensors by alternating resolutions and rescoring detections."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.15423/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"cited_work_retraction","ran_at":"2026-05-19T16:21:58.911783Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"citation_quote_validity","ran_at":"2026-05-19T15:50:40.754606Z","status":"completed","version":"0.1.0","findings_count":0},{"name":"doi_title_agreement","ran_at":"2026-05-19T15:31:17.785083Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T15:30:42.762899Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"claim_evidence","ran_at":"2026-05-19T14:21:54.139672Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T13:33:22.701953Z","status":"skipped","version":"1.0.0","findings_count":0}],"snapshot_sha256":"dc27aab67f5779495d37343b22b2db1cd84271c81e276179a306c6e055cdb9a3"},"references":{"count":70,"sample":[{"doi":"","year":2021,"title":"S. C. Mukhopadhyay, S. K. S. Tyagi, N. K. Suryadevara, V . Piuri, F. Scotti, and S. Zeadally, ‘‘Artificial intelligence-based sensors for next generation iot applications: A review,’’IEEE Sensors Jour","work_id":"5496f742-a5df-49d2-8961-291fa140d355","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"10.1007/s10462-022-10141-4","year":2022,"title":"W. Su, L. Li, F. Liu, M. He, and X. Liang, ‘‘Ai on the edge: a comprehensive review,’’Artif. Intell. Rev., vol. 55, no. 8, p. 6125–6183, Dec. 2022. [Online]. Available: https://doi.org/10.1007/s10462-","work_id":"725f0cdd-f1f3-4bfb-b1d9-f7f1c60e21b3","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2018,"title":"W. Y u, F. Liang, X. He, W. G. Hatcher, C. Lu, J. Lin, and X. Y ang, ‘‘A survey on the edge computing for the internet of things,’’IEEE Access, vol. 6, pp. 6900–6919, 2018","work_id":"569827dc-1a1f-4389-9b95-5f67fbaff3c5","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2022,"title":"K. S. Patle, R. Saini, A. Kumar, and V . S. Palaparthy, ‘‘Field evaluation of smart sensor system for plant disease prediction using lstm network,’’ IEEE Sensors Journal, vol. 22, no. 4, pp. 3715–3725","work_id":"0c21bad2-bcdf-4f7a-ad83-e35275c434d0","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2023,"title":"A. Sabato, S. Dabetwar, N. N. Kulkarni, and G. Fortino, ‘‘Noncontact sensing techniques for ai-aided structural health monitoring: A systematic review,’’IEEE Sensors Journal, vol. 23, no. 5, pp. 4672–","work_id":"b7740639-464a-4ad2-8e56-992b276b9230","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":70,"snapshot_sha256":"7ba3bd85a010425eb1b8c9d8d73464aeb01d85e22bd23eb46a9e5c50ea838de7","internal_anchors":7},"formal_canon":{"evidence_count":1,"snapshot_sha256":"e2c1c1298ab6d1159b1d591e913997ca449f043b96af8393d1df7bdd8dcc02bb"},"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":"2605.15423","created_at":"2026-05-20T00:00:57.844776+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.15423v1","created_at":"2026-05-20T00:00:57.844776+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.15423","created_at":"2026-05-20T00:00:57.844776+00:00"},{"alias_kind":"pith_short_12","alias_value":"4FRWERRYJFDT","created_at":"2026-05-20T00:00:57.844776+00:00"},{"alias_kind":"pith_short_16","alias_value":"4FRWERRYJFDTTKL6","created_at":"2026-05-20T00:00:57.844776+00:00"},{"alias_kind":"pith_short_8","alias_value":"4FRWERRY","created_at":"2026-05-20T00:00:57.844776+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":1,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/4FRWERRYJFDTTKL6ILZ24KR4LZ","json":"https://pith.science/pith/4FRWERRYJFDTTKL6ILZ24KR4LZ.json","graph_json":"https://pith.science/api/pith-number/4FRWERRYJFDTTKL6ILZ24KR4LZ/graph.json","events_json":"https://pith.science/api/pith-number/4FRWERRYJFDTTKL6ILZ24KR4LZ/events.json","paper":"https://pith.science/paper/4FRWERRY"},"agent_actions":{"view_html":"https://pith.science/pith/4FRWERRYJFDTTKL6ILZ24KR4LZ","download_json":"https://pith.science/pith/4FRWERRYJFDTTKL6ILZ24KR4LZ.json","view_paper":"https://pith.science/paper/4FRWERRY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.15423&json=true","fetch_graph":"https://pith.science/api/pith-number/4FRWERRYJFDTTKL6ILZ24KR4LZ/graph.json","fetch_events":"https://pith.science/api/pith-number/4FRWERRYJFDTTKL6ILZ24KR4LZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4FRWERRYJFDTTKL6ILZ24KR4LZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4FRWERRYJFDTTKL6ILZ24KR4LZ/action/storage_attestation","attest_author":"https://pith.science/pith/4FRWERRYJFDTTKL6ILZ24KR4LZ/action/author_attestation","sign_citation":"https://pith.science/pith/4FRWERRYJFDTTKL6ILZ24KR4LZ/action/citation_signature","submit_replication":"https://pith.science/pith/4FRWERRYJFDTTKL6ILZ24KR4LZ/action/replication_record"}},"created_at":"2026-05-20T00:00:57.844776+00:00","updated_at":"2026-05-20T00:00:57.844776+00:00"}