{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:SPHQANVJFYFZHFDZOA2ECYK4FX","short_pith_number":"pith:SPHQANVJ","schema_version":"1.0","canonical_sha256":"93cf0036a92e0b939479703441615c2dd3b06ebbf759f2f35228cdf74160979c","source":{"kind":"arxiv","id":"2603.18739","version":4},"attestation_state":"computed","paper":{"title":"EdgeCrafter: Compact ViTs for Edge Dense Prediction via Task-Specialized Distillation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Longfei Liu, Peizhe Ru, Qirui Wang, Xi Shen, Xuanlong Yu, Yang Li, Yinzhi Wang, Yongjie Hou, Yongjun Yu, Youyang Sha","submitted_at":"2026-03-19T10:39:51Z","abstract_excerpt":"Deploying high-performance dense prediction models on resource-constrained edge devices remains challenging due to strict computation and memory budgets. In practice, lightweight systems for object detection, instance segmentation, and pose estimation are still dominated by CNN-based architectures such as YOLO, while compact Vision Transformers (ViTs) often struggle to achieve comparable accuracy-efficiency trade-offs, even with large-scale pretraining. We argue that this gap arises primarily from insufficient task-specific representation learning in small-scale ViTs, rather than from an inher"},"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":"2603.18739","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-03-19T10:39:51Z","cross_cats_sorted":[],"title_canon_sha256":"a7e6b2c3ed592c43030dc2982c0582507aa749dd4ee44ddc6389a79756ef9938","abstract_canon_sha256":"baaee80e9711edd95a8ed41774047767a8dac53608c05f3b0221471490a8c974"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-18T01:14:19.732451Z","signature_b64":"DZMomhaUXK5EQcbhgdUXcG4g4f4HNLO303UCaRk4rgbEuUZ2CZKlFgjYRTz+Yl1oTjlevFkSVsPse//k9QUUBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"93cf0036a92e0b939479703441615c2dd3b06ebbf759f2f35228cdf74160979c","last_reissued_at":"2026-08-18T01:14:19.730638Z","signature_status":"signed_v1","first_computed_at":"2026-08-18T01:14:19.730638Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"EdgeCrafter: Compact ViTs for Edge Dense Prediction via Task-Specialized Distillation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Longfei Liu, Peizhe Ru, Qirui Wang, Xi Shen, Xuanlong Yu, Yang Li, Yinzhi Wang, Yongjie Hou, Yongjun Yu, Youyang Sha","submitted_at":"2026-03-19T10:39:51Z","abstract_excerpt":"Deploying high-performance dense prediction models on resource-constrained edge devices remains challenging due to strict computation and memory budgets. In practice, lightweight systems for object detection, instance segmentation, and pose estimation are still dominated by CNN-based architectures such as YOLO, while compact Vision Transformers (ViTs) often struggle to achieve comparable accuracy-efficiency trade-offs, even with large-scale pretraining. We argue that this gap arises primarily from insufficient task-specific representation learning in small-scale ViTs, rather than from an inher"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2603.18739","kind":"arxiv","version":4},"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/2603.18739/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":"2603.18739","created_at":"2026-08-18T01:14:19.731429+00:00"},{"alias_kind":"arxiv_version","alias_value":"2603.18739v4","created_at":"2026-08-18T01:14:19.731429+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2603.18739","created_at":"2026-08-18T01:14:19.731429+00:00"},{"alias_kind":"pith_short_12","alias_value":"SPHQANVJFYFZ","created_at":"2026-08-18T01:14:19.731429+00:00"},{"alias_kind":"pith_short_16","alias_value":"SPHQANVJFYFZHFDZ","created_at":"2026-08-18T01:14:19.731429+00:00"},{"alias_kind":"pith_short_8","alias_value":"SPHQANVJ","created_at":"2026-08-18T01:14:19.731429+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2608.03322","citing_title":"LocAnyMed: Vision-Language Grounding for Multimodal Medical Images","ref_index":45,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/SPHQANVJFYFZHFDZOA2ECYK4FX","json":"https://pith.science/pith/SPHQANVJFYFZHFDZOA2ECYK4FX.json","graph_json":"https://pith.science/api/pith-number/SPHQANVJFYFZHFDZOA2ECYK4FX/graph.json","events_json":"https://pith.science/api/pith-number/SPHQANVJFYFZHFDZOA2ECYK4FX/events.json","paper":"https://pith.science/paper/SPHQANVJ"},"agent_actions":{"view_html":"https://pith.science/pith/SPHQANVJFYFZHFDZOA2ECYK4FX","download_json":"https://pith.science/pith/SPHQANVJFYFZHFDZOA2ECYK4FX.json","view_paper":"https://pith.science/paper/SPHQANVJ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2603.18739&json=true","fetch_graph":"https://pith.science/api/pith-number/SPHQANVJFYFZHFDZOA2ECYK4FX/graph.json","fetch_events":"https://pith.science/api/pith-number/SPHQANVJFYFZHFDZOA2ECYK4FX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SPHQANVJFYFZHFDZOA2ECYK4FX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SPHQANVJFYFZHFDZOA2ECYK4FX/action/storage_attestation","attest_author":"https://pith.science/pith/SPHQANVJFYFZHFDZOA2ECYK4FX/action/author_attestation","sign_citation":"https://pith.science/pith/SPHQANVJFYFZHFDZOA2ECYK4FX/action/citation_signature","submit_replication":"https://pith.science/pith/SPHQANVJFYFZHFDZOA2ECYK4FX/action/replication_record"}},"created_at":"2026-08-18T01:14:19.731429+00:00","updated_at":"2026-08-18T01:14:19.731429+00:00"}