{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:VWDYEMZZPCKOTUXLCNC7LF3OIZ","short_pith_number":"pith:VWDYEMZZ","schema_version":"1.0","canonical_sha256":"ad878233397894e9d2eb1345f5976e4650decfda970d4b25f2b4b59126822681","source":{"kind":"arxiv","id":"2406.00670","version":2},"attestation_state":"computed","paper":{"title":"Cascade-CLIP: Cascaded Vision-Language Embeddings Alignment for Zero-Shot Semantic Segmentation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Ming-Ming Cheng, Qibin Hou, Quansheng Zeng, Yunheng Li, Zhongyu Li","submitted_at":"2024-06-02T08:32:51Z","abstract_excerpt":"Pre-trained vision-language models, e.g., CLIP, have been successfully applied to zero-shot semantic segmentation. Existing CLIP-based approaches primarily utilize visual features from the last layer to align with text embeddings, while they neglect the crucial information in intermediate layers that contain rich object details. However, we find that directly aggregating the multi-level visual features weakens the zero-shot ability for novel classes. The large differences between the visual features from different layers make these features hard to align well with the text embeddings. We resol"},"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":"2406.00670","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-06-02T08:32:51Z","cross_cats_sorted":[],"title_canon_sha256":"e33858a246950a26060b5cb525ff5f1c982ad106bc9e832b339c36d27a378341","abstract_canon_sha256":"75f4995c8ab166957678fd80dee05f398692c5b97c11a3d09d03b89708c930da"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:28:10.880712Z","signature_b64":"gZl0siKiCjBthbLc7q4EYWW6QjNWbaT0qJeOk0Zp/ObpW+GF4I54O4xkKNdbqj0OxfFn9AkrcHgk2S/wam1RAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ad878233397894e9d2eb1345f5976e4650decfda970d4b25f2b4b59126822681","last_reissued_at":"2026-07-05T08:28:10.880286Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:28:10.880286Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Cascade-CLIP: Cascaded Vision-Language Embeddings Alignment for Zero-Shot Semantic Segmentation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Ming-Ming Cheng, Qibin Hou, Quansheng Zeng, Yunheng Li, Zhongyu Li","submitted_at":"2024-06-02T08:32:51Z","abstract_excerpt":"Pre-trained vision-language models, e.g., CLIP, have been successfully applied to zero-shot semantic segmentation. Existing CLIP-based approaches primarily utilize visual features from the last layer to align with text embeddings, while they neglect the crucial information in intermediate layers that contain rich object details. However, we find that directly aggregating the multi-level visual features weakens the zero-shot ability for novel classes. The large differences between the visual features from different layers make these features hard to align well with the text embeddings. We resol"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.00670","kind":"arxiv","version":2},"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/2406.00670/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":"2406.00670","created_at":"2026-07-05T08:28:10.880343+00:00"},{"alias_kind":"arxiv_version","alias_value":"2406.00670v2","created_at":"2026-07-05T08:28:10.880343+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.00670","created_at":"2026-07-05T08:28:10.880343+00:00"},{"alias_kind":"pith_short_12","alias_value":"VWDYEMZZPCKO","created_at":"2026-07-05T08:28:10.880343+00:00"},{"alias_kind":"pith_short_16","alias_value":"VWDYEMZZPCKOTUXL","created_at":"2026-07-05T08:28:10.880343+00:00"},{"alias_kind":"pith_short_8","alias_value":"VWDYEMZZ","created_at":"2026-07-05T08:28:10.880343+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2506.01015","citing_title":"AuralSAM2: Enabling SAM2 Hear Through Pyramid Audio-Visual Feature Prompting","ref_index":26,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/VWDYEMZZPCKOTUXLCNC7LF3OIZ","json":"https://pith.science/pith/VWDYEMZZPCKOTUXLCNC7LF3OIZ.json","graph_json":"https://pith.science/api/pith-number/VWDYEMZZPCKOTUXLCNC7LF3OIZ/graph.json","events_json":"https://pith.science/api/pith-number/VWDYEMZZPCKOTUXLCNC7LF3OIZ/events.json","paper":"https://pith.science/paper/VWDYEMZZ"},"agent_actions":{"view_html":"https://pith.science/pith/VWDYEMZZPCKOTUXLCNC7LF3OIZ","download_json":"https://pith.science/pith/VWDYEMZZPCKOTUXLCNC7LF3OIZ.json","view_paper":"https://pith.science/paper/VWDYEMZZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2406.00670&json=true","fetch_graph":"https://pith.science/api/pith-number/VWDYEMZZPCKOTUXLCNC7LF3OIZ/graph.json","fetch_events":"https://pith.science/api/pith-number/VWDYEMZZPCKOTUXLCNC7LF3OIZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VWDYEMZZPCKOTUXLCNC7LF3OIZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VWDYEMZZPCKOTUXLCNC7LF3OIZ/action/storage_attestation","attest_author":"https://pith.science/pith/VWDYEMZZPCKOTUXLCNC7LF3OIZ/action/author_attestation","sign_citation":"https://pith.science/pith/VWDYEMZZPCKOTUXLCNC7LF3OIZ/action/citation_signature","submit_replication":"https://pith.science/pith/VWDYEMZZPCKOTUXLCNC7LF3OIZ/action/replication_record"}},"created_at":"2026-07-05T08:28:10.880343+00:00","updated_at":"2026-07-05T08:28:10.880343+00:00"}