{"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/coarse-to-fine-vision-language-pre-training","title":"Coarse-to-Fine Vision-Language Pre-training with Fusion in the Backbone","arxiv_id":"2206.07643","date":"2022-06-15","proceeding":"NeurIPS 2022 5","authors":["Zi-Yi Dou","Aishwarya Kamath","Zhe Gan","Pengchuan Zhang","JianFeng Wang","Linjie Li","Zicheng Liu","Ce Liu","Yann Lecun","Nanyun Peng","Jianfeng Gao","Lijuan Wang"],"abstract":"Vision-language (VL) pre-training has recently received considerable attention. However, most existing end-to-end pre-training approaches either only aim to tackle VL tasks such as image-text retrieval, visual question answering (VQA) and image captioning that test high-level understanding of images, or only target region-level understanding for tasks such as phrase grounding and object detection. We present FIBER (Fusion-In-the-Backbone-based transformER), a new VL model architecture that can seamlessly handle both these types of tasks. Instead of having dedicated transformer layers for fusion after the uni-modal backbones, FIBER pushes multimodal fusion deep into the model by inserting cross-attention into the image and text backbones, bringing gains in terms of memory and performance. In addition, unlike previous work that is either only pre-trained on image-text data or on fine-grained data with box-level annotations, we present a two-stage pre-training strategy that uses both these kinds of data efficiently: (i) coarse-grained pre-training based on image-text data; followed by (ii) fine-grained pre-training based on image-text-box data. We conduct comprehensive experiments on a wide range of VL tasks, ranging from VQA, image captioning, and retrieval, to phrase grounding, referring expression comprehension, and object detection. Using deep multimodal fusion coupled with the two-stage pre-training, FIBER provides consistent performance improvements over strong baselines across all tasks, often outperforming methods using magnitudes more data. Code is available at https://github.com/microsoft/FIBER.","url_abs":"https://arxiv.org/abs/2206.07643v2","url_pdf":"https://arxiv.org/pdf/2206.07643v2.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":"coarse-to-fine-vision-language-pre-training","repo_url":"https://github.com/microsoft/fiber","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"described-object-detection","task_name":"Described Object Detection"},{"task_slug":"image-captioning","task_name":"Image Captioning"},{"task_slug":"image-text-retrieval","task_name":"Image-text Retrieval"},{"task_slug":"object-detection","task_name":"Object Detection"},{"task_slug":"phrase-grounding","task_name":"Phrase Grounding"},{"task_slug":"question-answering","task_name":"Question Answering"},{"task_slug":"referring-expression-comprehension","task_name":"Referring Expression Comprehension"},{"task_slug":"text-retrieval","task_name":"Text Retrieval"},{"task_slug":"visual-question-answering-1","task_name":"Visual Question Answering"},{"task_slug":"visual-question-answering","task_name":"Visual Question Answering (VQA)"},{"task_slug":"visual-reasoning","task_name":"Visual Reasoning"}],"methods":[{"method_slug":"test","method_name":"Test"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/described-object-detection-on-description","task":"Described Object Detection","dataset":"Description Detection Dataset","model":"FIBER-B","rank_in_archive_order":2,"of":8,"metrics":{"Intra-scenario ABS mAP":"26.0","Intra-scenario FULL mAP":"22.7","Intra-scenario PRES mAP":"21.5"},"uses_additional_data":false},{"leaderboard":"/sota/object-detection-on-coco-o","task":"Object Detection","dataset":"COCO-O","model":"FIBER-B\n(Swin-B)","rank_in_archive_order":12,"of":45,"metrics":{"Average mAP":"33.7","Effective Robustness":"11.43"},"uses_additional_data":false},{"leaderboard":"/sota/phrase-grounding-on-flickr30k-entities-dev","task":"Phrase Grounding","dataset":"Flickr30k Entities Dev","model":"Fiber-B","rank_in_archive_order":1,"of":3,"metrics":{"R@1":"87.1","R@10":"97.4","R@5":"96.1"},"uses_additional_data":false},{"leaderboard":"/sota/phrase-grounding-on-flickr30k-entities-test","task":"Phrase Grounding","dataset":"Flickr30k Entities Test","model":"FIBER-B","rank_in_archive_order":2,"of":18,"metrics":{"R@1":"87.4","R@10":"97.6","R@5":"96.4"},"uses_additional_data":true}],"syntology":{"syntology_url":"https://syntology.ai/paper/2206.07643","atlas_url":"https://app.syntology.ai/?focus=2206.07643","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2206.07643"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-25T09:33:49+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/microsoft/FIBER","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/microsoft/fiber","reach":{"status":"ok"}}],"summary":{"ran_honours":1,"ran_draft_wrong":1,"unverified":1},"by_repo_kind":{"official":{"samples":1,"ran":0,"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":2,"samples":[{"code_sha256_prefix":"525b77c721203779","entry":"create_positive_map","repo":null,"repo_kind":null,"path":null,"file_url":null,"link_basis":"identical_code_first_harvested_elsewhere","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":"well_formed","behaviour_fingerprint":false,"licence":null,"inline_ok":false,"mcp_get_code":{"code_sha256":"525b77c721203779"}},{"code_sha256_prefix":"ea2c188d68558611","entry":"create_positive_map_label_to_token_from_positive_map","repo":null,"repo_kind":null,"path":null,"file_url":null,"link_basis":"identical_code_first_harvested_elsewhere","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":null,"inline_ok":false,"mcp_get_code":{"code_sha256":"ea2c188d68558611"}},{"code_sha256_prefix":"d772fb1ce96f80c5","entry":"FusionSwinTransformer","repo":"microsoft/fiber","repo_kind":"official","path":"fine_grained/maskrcnn_benchmark/modeling/backbone/fusion_swin_transformer.py","file_url":"https://github.com/microsoft/fiber/blob/HEAD/fine_grained/maskrcnn_benchmark/modeling/backbone/fusion_swin_transformer.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"d772fb1ce96f80c5"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}