{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/fine-grained-late-interaction-multi-modal-1","title":"Fine-grained Late-interaction Multi-modal Retrieval for Retrieval Augmented Visual Question Answering","arxiv_id":"2309.17133","date":"2023-09-29","proceeding":"NeurIPS 2023 11","authors":["Weizhe Lin","Jinghong Chen","Jingbiao Mei","Alexandru Coca","Bill Byrne"],"abstract":"Knowledge-based Visual Question Answering (KB-VQA) requires VQA systems to utilize knowledge from external knowledge bases to answer visually-grounded questions. Retrieval-Augmented Visual Question Answering (RA-VQA), a strong framework to tackle KB-VQA, first retrieves related documents with Dense Passage Retrieval (DPR) and then uses them to answer questions. This paper proposes Fine-grained Late-interaction Multi-modal Retrieval (FLMR) which significantly improves knowledge retrieval in RA-VQA. FLMR addresses two major limitations in RA-VQA's retriever: (1) the image representations obtained via image-to-text transforms can be incomplete and inaccurate and (2) relevance scores between queries and documents are computed with one-dimensional embeddings, which can be insensitive to finer-grained relevance. FLMR overcomes these limitations by obtaining image representations that complement those from the image-to-text transforms using a vision model aligned with an existing text-based retriever through a simple alignment network. FLMR also encodes images and questions using multi-dimensional embeddings to capture finer-grained relevance between queries and documents. FLMR significantly improves the original RA-VQA retriever's PRRecall@5 by approximately 8\\%. Finally, we equipped RA-VQA with two state-of-the-art large multi-modal/language models to achieve $\\sim61\\%$ VQA score in the OK-VQA dataset.","url_abs":"https://arxiv.org/abs/2309.17133v2","url_pdf":"https://arxiv.org/pdf/2309.17133v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"fine-grained-late-interaction-multi-modal-1","repo_url":"https://github.com/linweizhedragon/retrieval-augmented-visual-question-answering","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"GPL-3.0"}}],"tasks":[{"task_slug":"image-to-text","task_name":"Image to text"},{"task_slug":"passage-retrieval","task_name":"Passage Retrieval"},{"task_slug":"question-answering","task_name":"Question Answering"},{"task_slug":"retrieval","task_name":"Retrieval"},{"task_slug":"visual-question-answering-1","task_name":"Visual Question Answering"},{"task_slug":"visual-question-answering","task_name":"Visual Question Answering (VQA)"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/retrieval-on-ok-vqa","task":"Retrieval","dataset":"OK-VQA","model":"FLMR","rank_in_archive_order":1,"of":2,"metrics":{"Recall@5":"89.32"},"uses_additional_data":false},{"leaderboard":"/sota/visual-question-answering-on-ok-vqa","task":"Visual Question Answering (VQA)","dataset":"OK-VQA","model":"RA-VQA-v2 (BLIP 2)","rank_in_archive_order":6,"of":37,"metrics":{"Accuracy":"62.08","Exact Match (EM)":"62.01","Recall@5":"89.32"},"uses_additional_data":false},{"leaderboard":"/sota/visual-question-answering-on-ok-vqa","task":"Visual Question Answering (VQA)","dataset":"OK-VQA","model":"RA-VQA-v2 (T5-large)","rank_in_archive_order":13,"of":37,"metrics":{"Accuracy":"54.85"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2309.17133","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2309.17133"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"deterministic:regex_extraction","url":"https://github.com/LinWeizheDragon/Retrieval-Augmented-Visual-Question-Answering","reach":{"status":"ok","spdx":"GPL-3.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/linweizhedragon/retrieval-augmented-visual-question-answering","reach":{"status":"ok","spdx":"GPL-3.0"}}],"summary":{"ran_draft_wrong":5,"ran":3},"by_repo_kind":{"official":{"samples":8,"ran":8,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":8,"samples":[{"code_sha256_prefix":"9704017d6a608db7","entry":"compute_ranks","repo":"linweizhedragon/retrieval-augmented-visual-question-answering","repo_kind":"official","path":"materials/Oscar/oscar/run_retrieval.py","file_url":"https://github.com/linweizhedragon/retrieval-augmented-visual-question-answering/blob/HEAD/materials/Oscar/oscar/run_retrieval.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"GPL-3.0","inline_ok":false,"mcp_get_code":{"code_sha256":"9704017d6a608db7"}},{"code_sha256_prefix":"f34c3eff1c9e6c54","entry":"compute_score_with_logits","repo":"linweizhedragon/retrieval-augmented-visual-question-answering","repo_kind":"official","path":"materials/Oscar/oscar/run_retrieval.py","file_url":"https://github.com/linweizhedragon/retrieval-augmented-visual-question-answering/blob/HEAD/materials/Oscar/oscar/run_retrieval.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"GPL-3.0","inline_ok":false,"mcp_get_code":{"code_sha256":"f34c3eff1c9e6c54"}},{"code_sha256_prefix":"e36ec4e0a52c6a4b","entry":"instance_bce_with_logits","repo":"linweizhedragon/retrieval-augmented-visual-question-answering","repo_kind":"official","path":"materials/Oscar/oscar/run_vqa.py","file_url":"https://github.com/linweizhedragon/retrieval-augmented-visual-question-answering/blob/HEAD/materials/Oscar/oscar/run_vqa.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"invariant","behaviour_fingerprint":false,"licence":"GPL-3.0","inline_ok":false,"mcp_get_code":{"code_sha256":"e36ec4e0a52c6a4b"}},{"code_sha256_prefix":"de0d0e47ac6bbe6c","entry":"precision_at_k","repo":"LinWeizheDragon/Retrieval-Augmented-Visual-Question-Answering","repo_kind":"official","path":"src/utils/retrieval_metrics.py","file_url":"https://github.com/LinWeizheDragon/Retrieval-Augmented-Visual-Question-Answering/blob/HEAD/src/utils/retrieval_metrics.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"GPL-3.0","inline_ok":false,"mcp_get_code":{"code_sha256":"de0d0e47ac6bbe6c"}},{"code_sha256_prefix":"e06945a77a9e2b0a","entry":"precision_at_k_batch","repo":"LinWeizheDragon/Retrieval-Augmented-Visual-Question-Answering","repo_kind":"official","path":"src/utils/retrieval_metrics.py","file_url":"https://github.com/LinWeizheDragon/Retrieval-Augmented-Visual-Question-Answering/blob/HEAD/src/utils/retrieval_metrics.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"GPL-3.0","inline_ok":false,"mcp_get_code":{"code_sha256":"e06945a77a9e2b0a"}},{"code_sha256_prefix":"4437293713f250ee","entry":"recall","repo":"LinWeizheDragon/Retrieval-Augmented-Visual-Question-Answering","repo_kind":"official","path":"src/utils/retrieval_metrics.py","file_url":"https://github.com/LinWeizheDragon/Retrieval-Augmented-Visual-Question-Answering/blob/HEAD/src/utils/retrieval_metrics.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"GPL-3.0","inline_ok":false,"mcp_get_code":{"code_sha256":"4437293713f250ee"}},{"code_sha256_prefix":"9dfcf8b13847c65f","entry":"shift_tokens_right","repo":"LinWeizheDragon/Retrieval-Augmented-Visual-Question-Answering","repo_kind":"official","path":"src/models/custom_peft.py","file_url":"https://github.com/LinWeizheDragon/Retrieval-Augmented-Visual-Question-Answering/blob/HEAD/src/models/custom_peft.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":"GPL-3.0","inline_ok":false,"mcp_get_code":{"code_sha256":"9dfcf8b13847c65f"}},{"code_sha256_prefix":"12516cd21b436092","entry":"trim_batch","repo":"linweizhedragon/retrieval-augmented-visual-question-answering","repo_kind":"official","path":"materials/Oscar/oscar/run_vqa.py","file_url":"https://github.com/linweizhedragon/retrieval-augmented-visual-question-answering/blob/HEAD/materials/Oscar/oscar/run_vqa.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"GPL-3.0","inline_ok":false,"mcp_get_code":{"code_sha256":"12516cd21b436092"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}