{"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/how-much-can-clip-benefit-vision-and-language","title":"How Much Can CLIP Benefit Vision-and-Language Tasks?","arxiv_id":"2107.06383","date":"2021-07-13","proceeding":null,"authors":["Sheng Shen","Liunian Harold Li","Hao Tan","Mohit Bansal","Anna Rohrbach","Kai-Wei Chang","Zhewei Yao","Kurt Keutzer"],"abstract":"Most existing Vision-and-Language (V&L) models rely on pre-trained visual encoders, using a relatively small set of manually-annotated data (as compared to web-crawled data), to perceive the visual world. However, it has been observed that large-scale pretraining usually can result in better generalization performance, e.g., CLIP (Contrastive Language-Image Pre-training), trained on a massive amount of image-caption pairs, has shown a strong zero-shot capability on various vision tasks. To further study the advantage brought by CLIP, we propose to use CLIP as the visual encoder in various V&L models in two typical scenarios: 1) plugging CLIP into task-specific fine-tuning; 2) combining CLIP with V&L pre-training and transferring to downstream tasks. We show that CLIP significantly outperforms widely-used visual encoders trained with in-domain annotated data, such as BottomUp-TopDown. We achieve competitive or better results on diverse V&L tasks, while establishing new state-of-the-art results on Visual Question Answering, Visual Entailment, and V&L Navigation tasks. We release our code at https://github.com/clip-vil/CLIP-ViL.","url_abs":"https://arxiv.org/abs/2107.06383v1","url_pdf":"https://arxiv.org/pdf/2107.06383v1.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":"how-much-can-clip-benefit-vision-and-language","repo_url":"https://github.com/clip-vil/CLIP-ViL","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"how-much-can-clip-benefit-vision-and-language","repo_url":"https://github.com/facebookresearch/reliable_vqa","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"how-much-can-clip-benefit-vision-and-language","repo_url":"https://github.com/gchhablani/multilingual-vqa","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"jax","reach":null},{"paper_slug":"how-much-can-clip-benefit-vision-and-language","repo_url":"https://github.com/jianjieluo/openai-clip-feature","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"question-answering","task_name":"Question Answering"},{"task_slug":"vision-and-language-navigation","task_name":"Vision and Language Navigation"},{"task_slug":"visual-entailment","task_name":"Visual Entailment"},{"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/vision-and-language-navigation-on-rxr","task":"Vision and Language Navigation","dataset":"RxR","model":"CLEAR-CLIP","rank_in_archive_order":4,"of":6,"metrics":{"ndtw":"53.69"},"uses_additional_data":true},{"leaderboard":"/sota/visual-entailment-on-snli-ve-val","task":"Visual Entailment","dataset":"SNLI-VE val","model":"CLIP-ViL","rank_in_archive_order":6,"of":9,"metrics":{"Accuracy":"80.20"},"uses_additional_data":true}],"syntology":{"syntology_url":"https://syntology.ai/paper/2107.06383","atlas_url":"https://app.syntology.ai/?focus=2107.06383","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2107.06383"}},"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":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/clip-vil/CLIP-ViL","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/facebookresearch/reliable_vqa","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/jianjieluo/openai-clip-feature","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/gchhablani/multilingual-vqa","reach":null}],"summary":{"ran_draft_wrong":5,"ran":2,"unverified":3},"by_repo_kind":{"official":{"samples":1,"ran":0,"repositories":1},"listed":{"samples":9,"ran":7,"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":9,"samples":[{"code_sha256_prefix":"98f385d847636a3e","entry":"basic_clean","repo":"jianjieluo/openai-clip-feature","repo_kind":"listed","path":"clip/simple_tokenizer.py","file_url":"https://github.com/jianjieluo/openai-clip-feature/blob/HEAD/clip/simple_tokenizer.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":"MIT","inline_ok":false,"mcp_get_code":{"code_sha256":"98f385d847636a3e"}},{"code_sha256_prefix":"012bdd69e64bf2d1","entry":"custom_path2id","repo":"jianjieluo/openai-clip-feature","repo_kind":"listed","path":"clip_visual_feats.py","file_url":"https://github.com/jianjieluo/openai-clip-feature/blob/HEAD/clip_visual_feats.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":false,"mcp_get_code":{"code_sha256":"012bdd69e64bf2d1"}},{"code_sha256_prefix":"d919ae32e5e4e616","entry":"get_pairs","repo":"jianjieluo/openai-clip-feature","repo_kind":"listed","path":"clip/simple_tokenizer.py","file_url":"https://github.com/jianjieluo/openai-clip-feature/blob/HEAD/clip/simple_tokenizer.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"MIT","inline_ok":false,"mcp_get_code":{"code_sha256":"d919ae32e5e4e616"}},{"code_sha256_prefix":"2d946250dd2f5a4f","entry":"load_json","repo":"jianjieluo/openai-clip-feature","repo_kind":"listed","path":"basic_utils.py","file_url":"https://github.com/jianjieluo/openai-clip-feature/blob/HEAD/basic_utils.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"MIT","inline_ok":false,"mcp_get_code":{"code_sha256":"2d946250dd2f5a4f"}},{"code_sha256_prefix":"f1d1cccccf038785","entry":"load_jsonl","repo":"jianjieluo/openai-clip-feature","repo_kind":"listed","path":"basic_utils.py","file_url":"https://github.com/jianjieluo/openai-clip-feature/blob/HEAD/basic_utils.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":false,"mcp_get_code":{"code_sha256":"f1d1cccccf038785"}},{"code_sha256_prefix":"570ad34bd1af44a8","entry":"load_pickle","repo":"jianjieluo/openai-clip-feature","repo_kind":"listed","path":"basic_utils.py","file_url":"https://github.com/jianjieluo/openai-clip-feature/blob/HEAD/basic_utils.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"MIT","inline_ok":false,"mcp_get_code":{"code_sha256":"570ad34bd1af44a8"}},{"code_sha256_prefix":"9542161e9640b858","entry":"whitespace_clean","repo":"jianjieluo/openai-clip-feature","repo_kind":"listed","path":"clip/simple_tokenizer.py","file_url":"https://github.com/jianjieluo/openai-clip-feature/blob/HEAD/clip/simple_tokenizer.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":"MIT","inline_ok":false,"mcp_get_code":{"code_sha256":"9542161e9640b858"}},{"code_sha256_prefix":"0a59bad0d461993c","entry":"build_model","repo":"clip-vil/CLIP-ViL","repo_kind":"official","path":"CLIP-ViL-Direct/vqa/clip/model.py","file_url":"https://github.com/clip-vil/CLIP-ViL/blob/HEAD/CLIP-ViL-Direct/vqa/clip/model.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":"0a59bad0d461993c"}},{"code_sha256_prefix":"25a074ccf6e3f98a","entry":"build_model","repo":"jianjieluo/openai-clip-feature","repo_kind":"listed","path":"clip/model.py","file_url":"https://github.com/jianjieluo/openai-clip-feature/blob/HEAD/clip/model.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":false,"mcp_get_code":{"code_sha256":"25a074ccf6e3f98a"}},{"code_sha256_prefix":"2da7ec0975be872a","entry":"load","repo":"jianjieluo/openai-clip-feature","repo_kind":"listed","path":"clip/clip.py","file_url":"https://github.com/jianjieluo/openai-clip-feature/blob/HEAD/clip/clip.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":false,"mcp_get_code":{"code_sha256":"2da7ec0975be872a"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}