{"paper":{"title":"Text-Guided Visual Representation Learning for Robust Multimodal E-Commerce Recommendation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"TGQ-Former uses metadata as text guidance to extract robust visual tokens from cluttered product images for e-commerce retrieval.","cross_cats":[],"primary_cat":"cs.IR","authors_text":"Jing Ma, Jungong Han, Pinghua Gong, Shijie Yang, Tianlu Zhang, Weijie Ding, Yanlong Zang, Yufei Guo","submitted_at":"2026-05-17T10:20:23Z","abstract_excerpt":"Multimodal item embeddings are crucial for e-commerce item-to-item (I2I) retrieval, yet real-world product images often contain promotional overlays and background clutter that inject spurious visual cues and degrade retrieval robustness. This issue is particularly pronounced in MLRM-style pipelines, where a frozen vision encoder is connected to an LLM through a lightweight connector that must selectively aggregate visual tokens. We propose Text-Guided Q-Former (TGQ-Former), a text-guided visual representation learning framework that leverages structured metadata as semantic guidance for visua"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"TGQ-Former consistently outperforms strong connector baselines and end-to-end MLLMs on large-scale real-world e-commerce datasets with full-pool retrieval, improving Hit Rate@100 by 6.04% on average.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"Structured metadata is assumed to be accurate and sufficient to serve as reliable semantic guidance that allows the hybrid-query connector to disentangle metadata-anchored and exploratory visual streams without discarding useful visual evidence.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"TGQ-Former uses metadata-guided hybrid queries and dual-gated modulation to improve visual token selection in multimodal e-commerce retrieval, raising average Hit Rate@100 by 6.04% over baselines.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"TGQ-Former uses metadata as text guidance to extract robust visual tokens from cluttered product images for e-commerce retrieval.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"572b0de41057d05f0be1b78f17e871daaa05941cb2c71e141d8e8e11c2bc2fdc"},"source":{"id":"2605.17366","kind":"arxiv","version":1},"verdict":{"id":"84bfd2c5-0856-440c-8257-5f30a4e2dc34","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-19T23:04:45.034157Z","strongest_claim":"TGQ-Former consistently outperforms strong connector baselines and end-to-end MLLMs on large-scale real-world e-commerce datasets with full-pool retrieval, improving Hit Rate@100 by 6.04% on average.","one_line_summary":"TGQ-Former uses metadata-guided hybrid queries and dual-gated modulation to improve visual token selection in multimodal e-commerce retrieval, raising average Hit Rate@100 by 6.04% over baselines.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"Structured metadata is assumed to be accurate and sufficient to serve as reliable semantic guidance that allows the hybrid-query connector to disentangle metadata-anchored and exploratory visual streams without discarding useful visual evidence.","pith_extraction_headline":"TGQ-Former uses metadata as text guidance to extract robust visual tokens from cluttered product images for e-commerce retrieval."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.17366/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"doi_title_agreement","ran_at":"2026-05-19T23:31:20.074033Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T23:13:07.902413Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"claim_evidence","ran_at":"2026-05-19T21:41:57.781589Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T21:33:23.717398Z","status":"skipped","version":"1.0.0","findings_count":0}],"snapshot_sha256":"e232b7c4acd199a838180dc0b7124cba051141dfefc98fc4423e03bc58fd62d0"},"references":{"count":39,"sample":[{"doi":"","year":null,"title":"Jean-Baptiste Alayrac, Jeff Donahue, Pauline Luc, Antoine Miech, Iain Barr, Yana Hasson, Karel Lenc, Arthur Mensch, Katherine Millican, Malcolm Reynolds, et al","work_id":"fc034223-773d-41c1-9c67-26803f4f7b70","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2022,"title":"Flamingo: a visual language model for few-shot learning.Advances in neural information processing systems35 (2022), 23716–23736","work_id":"ff84ce1c-5fe0-420e-822f-cf901016e452","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"Qwen3-VL Technical Report","work_id":"1fe243aa-e3c0-4da6-b391-4cbcfc88d5c0","ref_index":3,"cited_arxiv_id":"2511.21631","is_internal_anchor":true},{"doi":"","year":2025,"title":"Qwen2.5-VL Technical Report","work_id":"69dffacb-bfe8-442d-be86-48624c60426f","ref_index":4,"cited_arxiv_id":"2502.13923","is_internal_anchor":true},{"doi":"10.1145/3331184.3331254","year":2019,"title":"Xu Chen, Hanxiong Chen, Hongteng Xu, Yongfeng Zhang, Yixin Cao, Zheng Qin, and Hongyuan Zha. 2019. Personalized Fashion Recommendation with Visual Explanations based on Multimodal Attention Network: T","work_id":"c6b0ed6f-ffae-4f38-a2df-ae8b2c9013a5","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":39,"snapshot_sha256":"931c9478ce21774350dd81aa692520a0241f75c75a56a94e310139e14f003f2d","internal_anchors":10},"formal_canon":{"evidence_count":2,"snapshot_sha256":"41629445618ad3180fcc85312e41d05e5880ec6e737e8009046d5127e89fc6a5"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}