{"about":{"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.","site":"https://codewithpapers.app","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","syntology":{"site":"https://syntology.ai","developers":"https://syntology.ai/developers","mcp":{"server":"https://syntology.ai/mcp","transport":"streamable-http","server_card":"https://syntology.ai/.well-known/mcp/server-card.json","auth":{"type":"trial token, no account","trial_token":"https://syntology.ai/api/oauth/trial/token","method":"POST","docs":"https://syntology.ai/developers"}},"have":"https://syntology.ai/api/graph/have?x=<method, arXiv id or title> (free, answers coverage only)","paper_base":"https://syntology.ai/paper/","atlas_base":"https://app.syntology.ai/?focus="},"machine_readable":[{"url":"https://codewithpapers.app/llms.txt","what":"the machine catalog: every machine-readable file, counted"},{"url":"https://codewithpapers.app/index/manifest.json","what":"paper-to-code index by arXiv id, with Syntology's counts"},{"url":"https://codewithpapers.app/search/manifest.json","what":"site search index (titles, authors) and its files"},{"url":"https://codewithpapers.app/download","what":"bulk files: Syntology's layer, described there"},{"url":"https://codewithpapers.app/build_manifest.json","what":"the build record: inputs, counts, exclusions, probes"}]},"url":"/paper/scaling-up-vision-language-pre-training-for","title":"Scaling Up Vision-Language Pre-training for Image Captioning","arxiv_id":"2111.12233","date":"2021-11-24","proceeding":"CVPR 2022 1","authors":["Xiaowei Hu","Zhe Gan","JianFeng Wang","Zhengyuan Yang","Zicheng Liu","Yumao Lu","Lijuan Wang"],"abstract":"In recent years, we have witnessed significant performance boost in the image captioning task based on vision-language pre-training (VLP). Scale is believed to be an important factor for this advance. However, most existing work only focuses on pre-training transformers with moderate sizes (e.g., 12 or 24 layers) on roughly 4 million images. In this paper, we present LEMON, a LargE-scale iMage captiONer, and provide the first empirical study on the scaling behavior of VLP for image captioning. We use the state-of-the-art VinVL model as our reference model, which consists of an image feature extractor and a transformer model, and scale the transformer both up and down, with model sizes ranging from 13 to 675 million parameters. In terms of data, we conduct experiments with up to 200 million image-text pairs which are automatically collected from web based on the alt attribute of the image (dubbed as ALT200M). Extensive analysis helps to characterize the performance trend as the model size and the pre-training data size increase. We also compare different training recipes, especially for training on large-scale noisy data. As a result, LEMON achieves new state of the arts on several major image captioning benchmarks, including COCO Caption, nocaps, and Conceptual Captions. We also show LEMON can generate captions with long-tail visual concepts when used in a zero-shot manner.","url_abs":"https://arxiv.org/abs/2111.12233v2","url_pdf":"https://arxiv.org/pdf/2111.12233v2.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":[],"tasks":[{"task_slug":"attribute","task_name":"Attribute"},{"task_slug":"image-captioning","task_name":"Image Captioning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/image-captioning-on-coco-captions","task":"Image Captioning","dataset":"COCO Captions","model":"LEMON","rank_in_archive_order":7,"of":41,"metrics":{"BLEU-4":"42.6","CIDER":"145.5","METEOR":"31.4","SPICE":"25.5"},"uses_additional_data":false},{"leaderboard":"/sota/image-captioning-on-nocaps-xd-entire","task":"Image Captioning","dataset":"nocaps-XD entire","model":"Microsoft Cognitive Services team","rank_in_archive_order":3,"of":12,"metrics":{"B1":"85.62","B2":"71.36","B3":"53.62","B4":"34.65","CIDEr":"114.25","METEOR":"31.27","ROUGE-L":"61.2","SPICE":"14.85"},"uses_additional_data":true},{"leaderboard":"/sota/image-captioning-on-nocaps-val-in-domain","task":"Image Captioning","dataset":"nocaps-val-in-domain","model":"LEMON_large","rank_in_archive_order":4,"of":11,"metrics":{"CIDEr":"116.9","Pre-train (#images)":" 200M","SPICE":"15.8"},"uses_additional_data":false},{"leaderboard":"/sota/image-captioning-on-nocaps-val-in-domain","task":"Image Captioning","dataset":"nocaps-val-in-domain","model":"LEMON_base","rank_in_archive_order":8,"of":11,"metrics":{"CIDEr":"107.7","Pre-train (#images)":"200M","SPICE":"14.7"},"uses_additional_data":false},{"leaderboard":"/sota/image-captioning-on-nocaps-val-near-domain","task":"Image Captioning","dataset":"nocaps-val-near-domain","model":"LEMON_large","rank_in_archive_order":4,"of":10,"metrics":{"CIDEr":"113.3","Pre-train (#images)":"200M","SPICE":" 15.1"},"uses_additional_data":false},{"leaderboard":"/sota/image-captioning-on-nocaps-val-out-domain","task":"Image Captioning","dataset":"nocaps-val-out-domain","model":"LEMON_large","rank_in_archive_order":7,"of":10,"metrics":{"CIDEr":"111.3","Pretrain (#images)":"200M","SPICE":" 14.0"},"uses_additional_data":false},{"leaderboard":"/sota/image-captioning-on-nocaps-val-overall","task":"Image Captioning","dataset":"nocaps-val-overall","model":"LEMON_large","rank_in_archive_order":4,"of":11,"metrics":{"CIDEr":"113.4","Pretrain (#images)":"200M","SPICE":" 15.0"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/2111.12233","atlas_url":"https://app.syntology.ai/?focus=2111.12233","mcp":null,"developers":"https://syntology.ai/developers","held":true,"reason":"arxiv_id_in_atlas","checked_against":{"atlas_ids":549449,"atlas_ids_sha256":"afd72399dafd9906e9062927f70655952385a3d75488558e1f0484e950ab09b7","read_at":"2026-10-05T01:46:21Z"},"agent_calls":[{"tool":"get_citation_path","arguments":{"paper_1":"2111.12233"},"arguments_in":{"paper_2":"another paper's arXiv id or title"},"call":"get_citation_path(paper_1=\"2111.12233\", paper_2=\"…\")"},{"tool":"get_concepts_for_paper","arguments":{"arxiv_id":"2111.12233"},"call":"get_concepts_for_paper(arxiv_id=\"2111.12233\")"}],"by_repo":[],"by_repo_is":"one row per repository in the archive's code links, Syntology's graph links or the samples; samples 0 means no sample is linked to this paper, a statement about Syntology's coverage, not about the repository: with harvested_for_other_papers present, Syntology harvested the repository for that many other papers, and without it, Syntology harvested nothing from it; placed_by_identical_code counts samples whose link records no repository, placed here because identical code was harvested from this repository for another paper; ran counts executions on a synthesized input and instrument counts failures of Syntology's instrument, not of the code"},"code_links_note":"the archive holds no code link for this paper","arxiv_metadata":null,"syntology_extracted_results":null,"record_sha256":"9795468e63bc385f8002f3d4bdf1e6cee301885fe95ce55e34a99d16149b48ea","record_changed_at":"2026-09-28","record_changed_at_basis":"first_hashed"}