{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:WUI4OANWUOZOLE7EESCBOXOGK5","short_pith_number":"pith:WUI4OANW","schema_version":"1.0","canonical_sha256":"b511c701b6a3b2e593e42484175dc6574bb9b6ed36825b3089309dc6a9683512","source":{"kind":"arxiv","id":"2411.18363","version":3},"attestation_state":"computed","paper":{"title":"ChatRex: Taming Multimodal LLM for Joint Perception and Understanding","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Gen Luo, Lei Zhang, Qing Jiang, Tianhe Ren, Yihao Chen, Yuda Xiong, Yuqin Yang, Zhaoyang Zeng","submitted_at":"2024-11-27T14:11:10Z","abstract_excerpt":"Perception and understanding are two pillars of computer vision. While multimodal large language models (MLLM) have demonstrated remarkable visual understanding capabilities, they arguably lack accurate perception abilities, e.g. the stage-of-the-art model Qwen2-VL only achieves a 43.9 recall rate on the COCO dataset, limiting many tasks requiring the combination of perception and understanding. In this work, we aim to bridge this perception gap from both model designing and data development perspectives. We first introduce ChatRex, an MLLM with a decoupled perception design. Instead of having"},"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":"2411.18363","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2024-11-27T14:11:10Z","cross_cats_sorted":[],"title_canon_sha256":"076e080982d13e378615af58c6662946a8f245ea9329e1b37735215dd79b5b90","abstract_canon_sha256":"a6c5205dfe31bb593c7c30a130dc5d713b5d46c41ea3255632aeff4a5f0df72d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:28:28.398002Z","signature_b64":"aZahYDzdpfzSMLQCRBhwj5gbJX1rOskyKnj9lHoD1ZdnEaYOV62Q6uPRybJsuRM8Yj+0OwaZv2jqKDtdwzQzBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b511c701b6a3b2e593e42484175dc6574bb9b6ed36825b3089309dc6a9683512","last_reissued_at":"2026-07-05T10:28:28.397509Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:28:28.397509Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ChatRex: Taming Multimodal LLM for Joint Perception and Understanding","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Gen Luo, Lei Zhang, Qing Jiang, Tianhe Ren, Yihao Chen, Yuda Xiong, Yuqin Yang, Zhaoyang Zeng","submitted_at":"2024-11-27T14:11:10Z","abstract_excerpt":"Perception and understanding are two pillars of computer vision. While multimodal large language models (MLLM) have demonstrated remarkable visual understanding capabilities, they arguably lack accurate perception abilities, e.g. the stage-of-the-art model Qwen2-VL only achieves a 43.9 recall rate on the COCO dataset, limiting many tasks requiring the combination of perception and understanding. In this work, we aim to bridge this perception gap from both model designing and data development perspectives. We first introduce ChatRex, an MLLM with a decoupled perception design. Instead of having"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.18363","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/2411.18363/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":"2411.18363","created_at":"2026-07-05T10:28:28.397573+00:00"},{"alias_kind":"arxiv_version","alias_value":"2411.18363v3","created_at":"2026-07-05T10:28:28.397573+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.18363","created_at":"2026-07-05T10:28:28.397573+00:00"},{"alias_kind":"pith_short_12","alias_value":"WUI4OANWUOZO","created_at":"2026-07-05T10:28:28.397573+00:00"},{"alias_kind":"pith_short_16","alias_value":"WUI4OANWUOZOLE7E","created_at":"2026-07-05T10:28:28.397573+00:00"},{"alias_kind":"pith_short_8","alias_value":"WUI4OANW","created_at":"2026-07-05T10:28:28.397573+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.06560","citing_title":"Vision as Unified Multimodal Generation","ref_index":77,"is_internal_anchor":true},{"citing_arxiv_id":"2606.26196","citing_title":"From Structure to Synergy: A Survey of Vision-Language Perception Paradigm Evolution in Multimodal Large Language Models","ref_index":28,"is_internal_anchor":false},{"citing_arxiv_id":"2605.14923","citing_title":"SceneParser: Hierarchical Scene Parsing for Visual Semantics Understanding","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2606.28862","citing_title":"HKVLM: Faithful Query--Region Binding for Frozen-Detector Visual Grounding","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2512.10554","citing_title":"Grounding Everything in Tokens for Multimodal Large Language Models","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2604.24036","citing_title":"Robust Grounding with MLLMs Against Occlusion and Small Objects via Language-Guided Semantic Cues","ref_index":13,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/WUI4OANWUOZOLE7EESCBOXOGK5","json":"https://pith.science/pith/WUI4OANWUOZOLE7EESCBOXOGK5.json","graph_json":"https://pith.science/api/pith-number/WUI4OANWUOZOLE7EESCBOXOGK5/graph.json","events_json":"https://pith.science/api/pith-number/WUI4OANWUOZOLE7EESCBOXOGK5/events.json","paper":"https://pith.science/paper/WUI4OANW"},"agent_actions":{"view_html":"https://pith.science/pith/WUI4OANWUOZOLE7EESCBOXOGK5","download_json":"https://pith.science/pith/WUI4OANWUOZOLE7EESCBOXOGK5.json","view_paper":"https://pith.science/paper/WUI4OANW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2411.18363&json=true","fetch_graph":"https://pith.science/api/pith-number/WUI4OANWUOZOLE7EESCBOXOGK5/graph.json","fetch_events":"https://pith.science/api/pith-number/WUI4OANWUOZOLE7EESCBOXOGK5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/WUI4OANWUOZOLE7EESCBOXOGK5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/WUI4OANWUOZOLE7EESCBOXOGK5/action/storage_attestation","attest_author":"https://pith.science/pith/WUI4OANWUOZOLE7EESCBOXOGK5/action/author_attestation","sign_citation":"https://pith.science/pith/WUI4OANWUOZOLE7EESCBOXOGK5/action/citation_signature","submit_replication":"https://pith.science/pith/WUI4OANWUOZOLE7EESCBOXOGK5/action/replication_record"}},"created_at":"2026-07-05T10:28:28.397573+00:00","updated_at":"2026-07-05T10:28:28.397573+00:00"}