{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:V7A74XFA77KUZGIMPFENEH7TXN","short_pith_number":"pith:V7A74XFA","schema_version":"1.0","canonical_sha256":"afc1fe5ca0ffd54c990c7948d21ff3bb62b31b25c6cf3120f284567c71c177aa","source":{"kind":"arxiv","id":"2501.15510","version":1},"attestation_state":"computed","paper":{"title":"Universal Image Restoration Pre-training via Degradation Classification","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Jiakui Hu, Lujia Jin, Yanye Lu, Zhengjian Yao","submitted_at":"2025-01-26T13:03:37Z","abstract_excerpt":"This paper proposes the Degradation Classification Pre-Training (DCPT), which enables models to learn how to classify the degradation type of input images for universal image restoration pre-training. Unlike the existing self-supervised pre-training methods, DCPT utilizes the degradation type of the input image as an extremely weak supervision, which can be effortlessly obtained, even intrinsic in all image restoration datasets. DCPT comprises two primary stages. Initially, image features are extracted from the encoder. Subsequently, a lightweight decoder, such as ResNet18, is leveraged to cla"},"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":"2501.15510","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-01-26T13:03:37Z","cross_cats_sorted":[],"title_canon_sha256":"7061b5916da14eb9df89b010d789494cd47be3e9f0b7274aeae9f1e199fcba58","abstract_canon_sha256":"dae1a24f6696531b535e07f0ac2c64e4498f3eb9db5f5c6bcb9124077e285c4a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:05:30.465802Z","signature_b64":"K/tG5wtCDhw4n4XKkSYSz1gMkAHz6JOH6/Sf3qHdCVSMGVh2l2yR5Efi8kRpWGGmClhvjEcNWCU+Yx6NP97DCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"afc1fe5ca0ffd54c990c7948d21ff3bb62b31b25c6cf3120f284567c71c177aa","last_reissued_at":"2026-07-05T10:05:30.465286Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:05:30.465286Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Universal Image Restoration Pre-training via Degradation Classification","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Jiakui Hu, Lujia Jin, Yanye Lu, Zhengjian Yao","submitted_at":"2025-01-26T13:03:37Z","abstract_excerpt":"This paper proposes the Degradation Classification Pre-Training (DCPT), which enables models to learn how to classify the degradation type of input images for universal image restoration pre-training. Unlike the existing self-supervised pre-training methods, DCPT utilizes the degradation type of the input image as an extremely weak supervision, which can be effortlessly obtained, even intrinsic in all image restoration datasets. DCPT comprises two primary stages. Initially, image features are extracted from the encoder. Subsequently, a lightweight decoder, such as ResNet18, is leveraged to cla"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.15510","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/2501.15510/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":"2501.15510","created_at":"2026-07-05T10:05:30.465354+00:00"},{"alias_kind":"arxiv_version","alias_value":"2501.15510v1","created_at":"2026-07-05T10:05:30.465354+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.15510","created_at":"2026-07-05T10:05:30.465354+00:00"},{"alias_kind":"pith_short_12","alias_value":"V7A74XFA77KU","created_at":"2026-07-05T10:05:30.465354+00:00"},{"alias_kind":"pith_short_16","alias_value":"V7A74XFA77KUZGIM","created_at":"2026-07-05T10:05:30.465354+00:00"},{"alias_kind":"pith_short_8","alias_value":"V7A74XFA","created_at":"2026-07-05T10:05:30.465354+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.19097","citing_title":"DVANet: Degradation-aware Visual-prior Alignment Network for Image Restoration","ref_index":55,"is_internal_anchor":false},{"citing_arxiv_id":"2601.20306","citing_title":"TPGDiff: Hierarchical Triple-Prior Guided Diffusion for Image Restoration","ref_index":28,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17506","citing_title":"Degradation Frequency Curve: An Explicit Frequency-Quantified Representation for All-in-One Image Restoration","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2604.09313","citing_title":"Compositional-Degradation UAV Image Restoration: Conditional Decoupled MoE Network and A Benchmark","ref_index":40,"is_internal_anchor":false},{"citing_arxiv_id":"2604.04924","citing_title":"Your Pre-trained Diffusion Model Secretly Knows Restoration","ref_index":19,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/V7A74XFA77KUZGIMPFENEH7TXN","json":"https://pith.science/pith/V7A74XFA77KUZGIMPFENEH7TXN.json","graph_json":"https://pith.science/api/pith-number/V7A74XFA77KUZGIMPFENEH7TXN/graph.json","events_json":"https://pith.science/api/pith-number/V7A74XFA77KUZGIMPFENEH7TXN/events.json","paper":"https://pith.science/paper/V7A74XFA"},"agent_actions":{"view_html":"https://pith.science/pith/V7A74XFA77KUZGIMPFENEH7TXN","download_json":"https://pith.science/pith/V7A74XFA77KUZGIMPFENEH7TXN.json","view_paper":"https://pith.science/paper/V7A74XFA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2501.15510&json=true","fetch_graph":"https://pith.science/api/pith-number/V7A74XFA77KUZGIMPFENEH7TXN/graph.json","fetch_events":"https://pith.science/api/pith-number/V7A74XFA77KUZGIMPFENEH7TXN/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/V7A74XFA77KUZGIMPFENEH7TXN/action/timestamp_anchor","attest_storage":"https://pith.science/pith/V7A74XFA77KUZGIMPFENEH7TXN/action/storage_attestation","attest_author":"https://pith.science/pith/V7A74XFA77KUZGIMPFENEH7TXN/action/author_attestation","sign_citation":"https://pith.science/pith/V7A74XFA77KUZGIMPFENEH7TXN/action/citation_signature","submit_replication":"https://pith.science/pith/V7A74XFA77KUZGIMPFENEH7TXN/action/replication_record"}},"created_at":"2026-07-05T10:05:30.465354+00:00","updated_at":"2026-07-05T10:05:30.465354+00:00"}