{"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/elevater-a-benchmark-and-toolkit-for","title":"ELEVATER: A Benchmark and Toolkit for Evaluating Language-Augmented Visual Models","arxiv_id":"2204.08790","date":"2022-04-19","proceeding":null,"authors":["Chunyuan Li","Haotian Liu","Liunian Harold Li","Pengchuan Zhang","Jyoti Aneja","Jianwei Yang","Ping Jin","Houdong Hu","Zicheng Liu","Yong Jae Lee","Jianfeng Gao"],"abstract":"Learning visual representations from natural language supervision has recently shown great promise in a number of pioneering works. In general, these language-augmented visual models demonstrate strong transferability to a variety of datasets and tasks. However, it remains challenging to evaluate the transferablity of these models due to the lack of easy-to-use evaluation toolkits and public benchmarks. To tackle this, we build ELEVATER (Evaluation of Language-augmented Visual Task-level Transfer), the first benchmark and toolkit for evaluating(pre-trained) language-augmented visual models. ELEVATER is composed of three components. (i) Datasets. As downstream evaluation suites, it consists of 20 image classification datasets and 35 object detection datasets, each of which is augmented with external knowledge. (ii) Toolkit. An automatic hyper-parameter tuning toolkit is developed to facilitate model evaluation on downstream tasks. (iii) Metrics. A variety of evaluation metrics are used to measure sample-efficiency (zero-shot and few-shot) and parameter-efficiency (linear probing and full model fine-tuning). ELEVATER is a platform for Computer Vision in the Wild (CVinW), and is publicly released at at https://computer-vision-in-the-wild.github.io/ELEVATER/","url_abs":"https://arxiv.org/abs/2204.08790v6","url_pdf":"https://arxiv.org/pdf/2204.08790v6.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":"elevater-a-benchmark-and-toolkit-for","repo_url":"https://github.com/Computer-Vision-in-the-Wild/Elevater_Toolkit_IC","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"elevater-a-benchmark-and-toolkit-for","repo_url":"https://github.com/computer-vision-in-the-wild/cvinw_readings","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}},{"paper_slug":"elevater-a-benchmark-and-toolkit-for","repo_url":"https://github.com/eric-ai-lab/pevit","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"elevater-a-benchmark-and-toolkit-for","repo_url":"https://github.com/microsoft/GLIP","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"elevater-a-benchmark-and-toolkit-for","repo_url":"https://github.com/microsoft/esvit","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"elevater-a-benchmark-and-toolkit-for","repo_url":"https://github.com/microsoft/klite","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"elevater-a-benchmark-and-toolkit-for","repo_url":"https://github.com/microsoft/unicl","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"elevater-a-benchmark-and-toolkit-for","repo_url":"https://github.com/rsCPSyEu/ovd_cod","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"elevater-a-benchmark-and-toolkit-for","repo_url":"https://github.com/sincerass/mvlpt","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"fairness","task_name":"Fairness"},{"task_slug":"few-shot-image-classification","task_name":"Few-Shot Image Classification"},{"task_slug":"few-shot-object-detection","task_name":"Few-Shot Object Detection"},{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"object-detection","task_name":"Object Detection"},{"task_slug":"zero-shot-image-classification","task_name":"Zero-Shot Image Classification"},{"task_slug":"zero-shot-object-detection","task_name":"Zero-Shot Object Detection"},{"task_slug":"image-classification","task_name":"image-classification"},{"task_slug":"object-detection-1","task_name":"object-detection"}],"methods":[],"datasets_introduced":[{"slug":"elevater","name":"ELEVATER","full_name":"Evaluation of Language-augmented Visual Task-level Transfer"}],"methods_introduced":[],"results":[{"leaderboard":"/sota/object-detection-on-elevater","task":"Object Detection","dataset":"ELEVATER","model":"GLIP-T","rank_in_archive_order":1,"of":1,"metrics":{"AP":"62.6"},"uses_additional_data":false},{"leaderboard":"/sota/object-detection-on-odinw-full-shot-35-tasks","task":"Object Detection","dataset":"ODinW Full-shot 35 Tasks","model":"GLIP-T","rank_in_archive_order":2,"of":2,"metrics":{"AP":"62.6"},"uses_additional_data":true},{"leaderboard":"/sota/zero-shot-image-classification-on-icinw","task":"Zero-Shot Image Classification","dataset":"ICinW","model":"CLIP (ViT B-32)","rank_in_archive_order":1,"of":1,"metrics":{"Average Score":"56.64"},"uses_additional_data":false},{"leaderboard":"/sota/zero-shot-image-classification-on-odinw","task":"Zero-Shot Image Classification","dataset":"ODinW","model":"GLIP (Tiny A)","rank_in_archive_order":1,"of":1,"metrics":{"Average Score":"11.4"},"uses_additional_data":false},{"leaderboard":"/sota/zero-shot-object-detection-on-odinw","task":"Zero-Shot Object Detection","dataset":"ODinW","model":"GLIP (Tiny A)","rank_in_archive_order":5,"of":5,"metrics":{"Average Score":"11.4"},"uses_additional_data":true}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2204.08790","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2204.08790"}},"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":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/microsoft/klite","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/microsoft/GLIP","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/sincerass/mvlpt","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/microsoft/unicl","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/Computer-Vision-in-the-Wild/Elevater_Toolkit_IC","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/microsoft/esvit","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/computer-vision-in-the-wild/cvinw_readings","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/rsCPSyEu/ovd_cod","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/eric-ai-lab/pevit","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran_fixture":2,"ran_draft_wrong":2,"unverified":16},"by_repo_kind":{"official":{"samples":11,"ran":2,"repositories":1},"listed":{"samples":9,"ran":2,"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":0,"samples":[{"code_sha256_prefix":"144d10b49baeb8a6","entry":"window_partition","repo":"Computer-Vision-in-the-Wild/Elevater_Toolkit_IC","repo_kind":"official","path":"vision_benchmark/models/cls_swin.py","file_url":"https://github.com/Computer-Vision-in-the-Wild/Elevater_Toolkit_IC/blob/HEAD/vision_benchmark/models/cls_swin.py","link_basis":"harvester_set","language":"python","status":"ran_fixture","verification_level":2,"contract_check":"RAISES","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"144d10b49baeb8a6"}},{"code_sha256_prefix":"b2ee26e229353a37","entry":"glorot_uniform","repo":"eric-ai-lab/pevit","repo_kind":"listed","path":"vision_benchmark/evaluation/compacter_model.py","file_url":"https://github.com/eric-ai-lab/pevit/blob/HEAD/vision_benchmark/evaluation/compacter_model.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"b2ee26e229353a37"}},{"code_sha256_prefix":"59e2991d3f405758","entry":"kronecker_product_einsum_batched","repo":"eric-ai-lab/pevit","repo_kind":"listed","path":"vision_benchmark/evaluation/compacter_model.py","file_url":"https://github.com/eric-ai-lab/pevit/blob/HEAD/vision_benchmark/evaluation/compacter_model.py","link_basis":"harvester_set","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"59e2991d3f405758"}},{"code_sha256_prefix":"61bf152e6a42a184","entry":"window_reverse","repo":"Computer-Vision-in-the-Wild/Elevater_Toolkit_IC","repo_kind":"official","path":"vision_benchmark/models/cls_swin.py","file_url":"https://github.com/Computer-Vision-in-the-Wild/Elevater_Toolkit_IC/blob/HEAD/vision_benchmark/models/cls_swin.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"invariant","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"61bf152e6a42a184"}},{"code_sha256_prefix":"ef2f936b8e5c816f","entry":"ask_gpt3","repo":"Computer-Vision-in-the-Wild/Elevater_Toolkit_IC","repo_kind":"official","path":"vision_benchmark/commands/extract_gpt3_knowledge.py","file_url":"https://github.com/Computer-Vision-in-the-Wild/Elevater_Toolkit_IC/blob/HEAD/vision_benchmark/commands/extract_gpt3_knowledge.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":true,"mcp_get_code":{"code_sha256":"ef2f936b8e5c816f"}},{"code_sha256_prefix":"497088982a2f6f81","entry":"build_adapter_model","repo":"eric-ai-lab/pevit","repo_kind":"listed","path":"vision_benchmark/evaluation/adapter_model.py","file_url":"https://github.com/eric-ai-lab/pevit/blob/HEAD/vision_benchmark/evaluation/adapter_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":true,"mcp_get_code":{"code_sha256":"497088982a2f6f81"}},{"code_sha256_prefix":"052dba32d0be4f7e","entry":"build_compacter_model","repo":"eric-ai-lab/pevit","repo_kind":"listed","path":"vision_benchmark/evaluation/compacter_model.py","file_url":"https://github.com/eric-ai-lab/pevit/blob/HEAD/vision_benchmark/evaluation/compacter_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":true,"mcp_get_code":{"code_sha256":"052dba32d0be4f7e"}},{"code_sha256_prefix":"ea0010b94502a91c","entry":"clone_loader","repo":"Computer-Vision-in-the-Wild/Elevater_Toolkit_IC","repo_kind":"official","path":"vision_benchmark/evaluation/full_model_finetune.py","file_url":"https://github.com/Computer-Vision-in-the-Wild/Elevater_Toolkit_IC/blob/HEAD/vision_benchmark/evaluation/full_model_finetune.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":true,"mcp_get_code":{"code_sha256":"ea0010b94502a91c"}},{"code_sha256_prefix":"0fcad17cd3cdfb8d","entry":"extract_finetune_results","repo":"eric-ai-lab/pevit","repo_kind":"listed","path":"read_results.py","file_url":"https://github.com/eric-ai-lab/pevit/blob/HEAD/read_results.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":true,"mcp_get_code":{"code_sha256":"0fcad17cd3cdfb8d"}},{"code_sha256_prefix":"d276b7d04501c57c","entry":"extract_finetune_results","repo":"eric-ai-lab/pevit","repo_kind":"listed","path":"read_txt.py","file_url":"https://github.com/eric-ai-lab/pevit/blob/HEAD/read_txt.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":true,"mcp_get_code":{"code_sha256":"d276b7d04501c57c"}},{"code_sha256_prefix":"ced21f6a4b3113c2","entry":"extract_finetune_results_para","repo":"eric-ai-lab/pevit","repo_kind":"listed","path":"read_txt.py","file_url":"https://github.com/eric-ai-lab/pevit/blob/HEAD/read_txt.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":true,"mcp_get_code":{"code_sha256":"ced21f6a4b3113c2"}},{"code_sha256_prefix":"aa72222fcaba5411","entry":"get_cls_model","repo":"Computer-Vision-in-the-Wild/Elevater_Toolkit_IC","repo_kind":"official","path":"vision_benchmark/models/cls_example.py","file_url":"https://github.com/Computer-Vision-in-the-Wild/Elevater_Toolkit_IC/blob/HEAD/vision_benchmark/models/cls_example.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":true,"mcp_get_code":{"code_sha256":"aa72222fcaba5411"}},{"code_sha256_prefix":"68f56d4679dbb36f","entry":"get_cls_model","repo":"Computer-Vision-in-the-Wild/Elevater_Toolkit_IC","repo_kind":"official","path":"vision_benchmark/models/cls_swin.py","file_url":"https://github.com/Computer-Vision-in-the-Wild/Elevater_Toolkit_IC/blob/HEAD/vision_benchmark/models/cls_swin.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":true,"mcp_get_code":{"code_sha256":"68f56d4679dbb36f"}},{"code_sha256_prefix":"aead1abefc725217","entry":"get_model","repo":"Computer-Vision-in-the-Wild/Elevater_Toolkit_IC","repo_kind":"official","path":"vision_benchmark/models/declip.py","file_url":"https://github.com/Computer-Vision-in-the-Wild/Elevater_Toolkit_IC/blob/HEAD/vision_benchmark/models/declip.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":true,"mcp_get_code":{"code_sha256":"aead1abefc725217"}},{"code_sha256_prefix":"853c0a124b0d326e","entry":"get_model","repo":"Computer-Vision-in-the-Wild/Elevater_Toolkit_IC","repo_kind":"official","path":"vision_benchmark/models/mae.py","file_url":"https://github.com/Computer-Vision-in-the-Wild/Elevater_Toolkit_IC/blob/HEAD/vision_benchmark/models/mae.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":true,"mcp_get_code":{"code_sha256":"853c0a124b0d326e"}},{"code_sha256_prefix":"c6f5f3a48a3d847d","entry":"get_model","repo":"Computer-Vision-in-the-Wild/Elevater_Toolkit_IC","repo_kind":"official","path":"vision_benchmark/models/mocov3.py","file_url":"https://github.com/Computer-Vision-in-the-Wild/Elevater_Toolkit_IC/blob/HEAD/vision_benchmark/models/mocov3.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":true,"mcp_get_code":{"code_sha256":"c6f5f3a48a3d847d"}},{"code_sha256_prefix":"deeacd37b9b6b8a5","entry":"get_zeroshot_model","repo":"Computer-Vision-in-the-Wild/Elevater_Toolkit_IC","repo_kind":"official","path":"vision_benchmark/models/clip_example.py","file_url":"https://github.com/Computer-Vision-in-the-Wild/Elevater_Toolkit_IC/blob/HEAD/vision_benchmark/models/clip_example.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":true,"mcp_get_code":{"code_sha256":"deeacd37b9b6b8a5"}},{"code_sha256_prefix":"c35a97c3423834d4","entry":"random_query","repo":"Computer-Vision-in-the-Wild/Elevater_Toolkit_IC","repo_kind":"official","path":"vision_benchmark/commands/extract_gpt3_knowledge.py","file_url":"https://github.com/Computer-Vision-in-the-Wild/Elevater_Toolkit_IC/blob/HEAD/vision_benchmark/commands/extract_gpt3_knowledge.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":true,"mcp_get_code":{"code_sha256":"c35a97c3423834d4"}},{"code_sha256_prefix":"4e6ec58f35d0977b","entry":"read_json","repo":"eric-ai-lab/pevit","repo_kind":"listed","path":"read_results.py","file_url":"https://github.com/eric-ai-lab/pevit/blob/HEAD/read_results.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":true,"mcp_get_code":{"code_sha256":"4e6ec58f35d0977b"}},{"code_sha256_prefix":"5f28aa08fa3d16f1","entry":"read_json","repo":"eric-ai-lab/pevit","repo_kind":"listed","path":"read_txt.py","file_url":"https://github.com/eric-ai-lab/pevit/blob/HEAD/read_txt.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":true,"mcp_get_code":{"code_sha256":"5f28aa08fa3d16f1"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}