{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:3TOIJZDAUD4JKQF5QKHRFF2PL2","short_pith_number":"pith:3TOIJZDA","schema_version":"1.0","canonical_sha256":"dcdc84e460a0f89540bd828f12974f5e87bbbb5110e4fb5c8dabe0f7bffbe77a","source":{"kind":"arxiv","id":"2412.11464","version":3},"attestation_state":"computed","paper":{"title":"High-Quality Mask Tuning Matters for Open-Vocabulary Segmentation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Daquan Zhou, Guanbin Li, Ming-Ming Cheng, Qibin Hou, Quan-Sheng Zeng, Yunheng Li","submitted_at":"2024-12-16T05:44:45Z","abstract_excerpt":"Open-vocabulary image segmentation has been advanced through the synergy between mask generators and vision-language models like Contrastive Language-Image Pre-training (CLIP). Previous approaches focus on generating masks while aligning mask features with text embeddings during training. In this paper, we observe that relying on generated low-quality masks can weaken the alignment of vision and language in regional representations. This motivates us to present a new fine-tuning framework, named MaskCLIP++, which uses ground-truth masks instead of generated masks to enhance the mask classifica"},"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":"2412.11464","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-12-16T05:44:45Z","cross_cats_sorted":[],"title_canon_sha256":"e1798b39cf8a41355469b277bab5d2a9a3db890be39a674b9caa86af6fd0d4b8","abstract_canon_sha256":"db284165b02d8b554a246998dae31d59f3ee3908cc13dd347d578d12ba73c11e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:29:30.981372Z","signature_b64":"qTre68JUCbjQUA9mJIXuvNHnO2QkxXjd/7O+sdx0EMsSVleblclZCorNdEnjdQeISNsyfdBkqQnFvJrnUqq7CQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dcdc84e460a0f89540bd828f12974f5e87bbbb5110e4fb5c8dabe0f7bffbe77a","last_reissued_at":"2026-07-05T10:29:30.980788Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:29:30.980788Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"High-Quality Mask Tuning Matters for Open-Vocabulary Segmentation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Daquan Zhou, Guanbin Li, Ming-Ming Cheng, Qibin Hou, Quan-Sheng Zeng, Yunheng Li","submitted_at":"2024-12-16T05:44:45Z","abstract_excerpt":"Open-vocabulary image segmentation has been advanced through the synergy between mask generators and vision-language models like Contrastive Language-Image Pre-training (CLIP). Previous approaches focus on generating masks while aligning mask features with text embeddings during training. In this paper, we observe that relying on generated low-quality masks can weaken the alignment of vision and language in regional representations. This motivates us to present a new fine-tuning framework, named MaskCLIP++, which uses ground-truth masks instead of generated masks to enhance the mask classifica"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.11464","kind":"arxiv","version":3},"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/2412.11464/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":"2412.11464","created_at":"2026-07-05T10:29:30.980851+00:00"},{"alias_kind":"arxiv_version","alias_value":"2412.11464v3","created_at":"2026-07-05T10:29:30.980851+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.11464","created_at":"2026-07-05T10:29:30.980851+00:00"},{"alias_kind":"pith_short_12","alias_value":"3TOIJZDAUD4J","created_at":"2026-07-05T10:29:30.980851+00:00"},{"alias_kind":"pith_short_16","alias_value":"3TOIJZDAUD4JKQF5","created_at":"2026-07-05T10:29:30.980851+00:00"},{"alias_kind":"pith_short_8","alias_value":"3TOIJZDA","created_at":"2026-07-05T10:29:30.980851+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.08001","citing_title":"Learning a Semantic Calibration Network for Open-Vocabulary Semantic Segmentation","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2605.19410","citing_title":"Vision Harnessing Agent for Open Ad-hoc Segmentation","ref_index":26,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3TOIJZDAUD4JKQF5QKHRFF2PL2","json":"https://pith.science/pith/3TOIJZDAUD4JKQF5QKHRFF2PL2.json","graph_json":"https://pith.science/api/pith-number/3TOIJZDAUD4JKQF5QKHRFF2PL2/graph.json","events_json":"https://pith.science/api/pith-number/3TOIJZDAUD4JKQF5QKHRFF2PL2/events.json","paper":"https://pith.science/paper/3TOIJZDA"},"agent_actions":{"view_html":"https://pith.science/pith/3TOIJZDAUD4JKQF5QKHRFF2PL2","download_json":"https://pith.science/pith/3TOIJZDAUD4JKQF5QKHRFF2PL2.json","view_paper":"https://pith.science/paper/3TOIJZDA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2412.11464&json=true","fetch_graph":"https://pith.science/api/pith-number/3TOIJZDAUD4JKQF5QKHRFF2PL2/graph.json","fetch_events":"https://pith.science/api/pith-number/3TOIJZDAUD4JKQF5QKHRFF2PL2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3TOIJZDAUD4JKQF5QKHRFF2PL2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3TOIJZDAUD4JKQF5QKHRFF2PL2/action/storage_attestation","attest_author":"https://pith.science/pith/3TOIJZDAUD4JKQF5QKHRFF2PL2/action/author_attestation","sign_citation":"https://pith.science/pith/3TOIJZDAUD4JKQF5QKHRFF2PL2/action/citation_signature","submit_replication":"https://pith.science/pith/3TOIJZDAUD4JKQF5QKHRFF2PL2/action/replication_record"}},"created_at":"2026-07-05T10:29:30.980851+00:00","updated_at":"2026-07-05T10:29:30.980851+00:00"}