{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:5THRQD65CZLHVOBZHEKVVMI2B5","short_pith_number":"pith:5THRQD65","schema_version":"1.0","canonical_sha256":"eccf180fdd16567ab83939155ab11a0f4014182d7952e917c36c0d6df24cd261","source":{"kind":"arxiv","id":"2405.14294","version":1},"attestation_state":"computed","paper":{"title":"Tuning-free Universally-Supervised Semantic Segmentation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Xiaobo Yang, Xiaojin Gong","submitted_at":"2024-05-23T08:13:52Z","abstract_excerpt":"This work presents a tuning-free semantic segmentation framework based on classifying SAM masks by CLIP, which is universally applicable to various types of supervision. Initially, we utilize CLIP's zero-shot classification ability to generate pseudo-labels or perform open-vocabulary segmentation. However, the misalignment between mask and CLIP text embeddings leads to suboptimal results. To address this issue, we propose discrimination-bias aligned CLIP to closely align mask and text embedding, offering an overhead-free performance gain. We then construct a global-local consistent classifier "},"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":"2405.14294","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-05-23T08:13:52Z","cross_cats_sorted":[],"title_canon_sha256":"3b3fb09b4917660d80ac35cc7f85242b83a97a365b5f02dce2ac2455569e9858","abstract_canon_sha256":"7d9ec981bdd2a85fe47698725fc989b3ca444a0ee772033b4f96b67b6dca86d7"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:22:12.499857Z","signature_b64":"MdgJ3NN05iiKxE+tmCAnhFdwLLHJ50pTp1PKVyId1uZRPuoIoF0b0RLhpI4EOUA5Nx5rUQ5VuiUC4uhCsGfGDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"eccf180fdd16567ab83939155ab11a0f4014182d7952e917c36c0d6df24cd261","last_reissued_at":"2026-07-05T08:22:12.499415Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:22:12.499415Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Tuning-free Universally-Supervised Semantic Segmentation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Xiaobo Yang, Xiaojin Gong","submitted_at":"2024-05-23T08:13:52Z","abstract_excerpt":"This work presents a tuning-free semantic segmentation framework based on classifying SAM masks by CLIP, which is universally applicable to various types of supervision. Initially, we utilize CLIP's zero-shot classification ability to generate pseudo-labels or perform open-vocabulary segmentation. However, the misalignment between mask and CLIP text embeddings leads to suboptimal results. To address this issue, we propose discrimination-bias aligned CLIP to closely align mask and text embedding, offering an overhead-free performance gain. We then construct a global-local consistent classifier "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.14294","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/2405.14294/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":"2405.14294","created_at":"2026-07-05T08:22:12.499472+00:00"},{"alias_kind":"arxiv_version","alias_value":"2405.14294v1","created_at":"2026-07-05T08:22:12.499472+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.14294","created_at":"2026-07-05T08:22:12.499472+00:00"},{"alias_kind":"pith_short_12","alias_value":"5THRQD65CZLH","created_at":"2026-07-05T08:22:12.499472+00:00"},{"alias_kind":"pith_short_16","alias_value":"5THRQD65CZLHVOBZ","created_at":"2026-07-05T08:22:12.499472+00:00"},{"alias_kind":"pith_short_8","alias_value":"5THRQD65","created_at":"2026-07-05T08:22:12.499472+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2411.15851","citing_title":"ResCLIP: Residual Attention for Training-free Dense Vision-language Inference","ref_index":54,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/5THRQD65CZLHVOBZHEKVVMI2B5","json":"https://pith.science/pith/5THRQD65CZLHVOBZHEKVVMI2B5.json","graph_json":"https://pith.science/api/pith-number/5THRQD65CZLHVOBZHEKVVMI2B5/graph.json","events_json":"https://pith.science/api/pith-number/5THRQD65CZLHVOBZHEKVVMI2B5/events.json","paper":"https://pith.science/paper/5THRQD65"},"agent_actions":{"view_html":"https://pith.science/pith/5THRQD65CZLHVOBZHEKVVMI2B5","download_json":"https://pith.science/pith/5THRQD65CZLHVOBZHEKVVMI2B5.json","view_paper":"https://pith.science/paper/5THRQD65","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2405.14294&json=true","fetch_graph":"https://pith.science/api/pith-number/5THRQD65CZLHVOBZHEKVVMI2B5/graph.json","fetch_events":"https://pith.science/api/pith-number/5THRQD65CZLHVOBZHEKVVMI2B5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5THRQD65CZLHVOBZHEKVVMI2B5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5THRQD65CZLHVOBZHEKVVMI2B5/action/storage_attestation","attest_author":"https://pith.science/pith/5THRQD65CZLHVOBZHEKVVMI2B5/action/author_attestation","sign_citation":"https://pith.science/pith/5THRQD65CZLHVOBZHEKVVMI2B5/action/citation_signature","submit_replication":"https://pith.science/pith/5THRQD65CZLHVOBZHEKVVMI2B5/action/replication_record"}},"created_at":"2026-07-05T08:22:12.499472+00:00","updated_at":"2026-07-05T08:22:12.499472+00:00"}