{"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/match-expand-and-improve-unsupervised","title":"MAtch, eXpand and Improve: Unsupervised Finetuning for Zero-Shot Action Recognition with Language Knowledge","arxiv_id":"2303.08914","date":"2023-03-15","proceeding":"ICCV 2023 1","authors":["Wei Lin","Leonid Karlinsky","Nina Shvetsova","Horst Possegger","Mateusz Kozinski","Rameswar Panda","Rogerio Feris","Hilde Kuehne","Horst Bischof"],"abstract":"Large scale Vision-Language (VL) models have shown tremendous success in aligning representations between visual and text modalities. This enables remarkable progress in zero-shot recognition, image generation & editing, and many other exciting tasks. However, VL models tend to over-represent objects while paying much less attention to verbs, and require additional tuning on video data for best zero-shot action recognition performance. While previous work relied on large-scale, fully-annotated data, in this work we propose an unsupervised approach. We adapt a VL model for zero-shot and few-shot action recognition using a collection of unlabeled videos and an unpaired action dictionary. Based on that, we leverage Large Language Models and VL models to build a text bag for each unlabeled video via matching, text expansion and captioning. We use those bags in a Multiple Instance Learning setup to adapt an image-text backbone to video data. Although finetuned on unlabeled video data, our resulting models demonstrate high transferability to numerous unseen zero-shot downstream tasks, improving the base VL model performance by up to 14\\%, and even comparing favorably to fully-supervised baselines in both zero-shot and few-shot video recognition transfer. The code will be released later at \\url{https://github.com/wlin-at/MAXI}.","url_abs":"https://arxiv.org/abs/2303.08914v2","url_pdf":"https://arxiv.org/pdf/2303.08914v2.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":"match-expand-and-improve-unsupervised","repo_url":"https://github.com/wlin-at/maxi","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"action-recognition-in-videos","task_name":"Action Recognition"},{"task_slug":"few-shot-action-recognition","task_name":"Few Shot Action Recognition"},{"task_slug":"few-shot-action-recognition","task_name":"Few-Shot action recognition"},{"task_slug":"image-generation","task_name":"Image Generation"},{"task_slug":"multiple-instance-learning","task_name":"Multiple Instance Learning"},{"task_slug":"video-recognition","task_name":"Video Recognition"},{"task_slug":"zero-shot-action-recognition","task_name":"Zero-Shot Action Recognition"},{"task_slug":"zero-shot-learning","task_name":"Zero-Shot Learning"}],"methods":[{"method_slug":"base","method_name":"BASE"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/zero-shot-action-recognition-on-charades-1","task":"Zero-Shot Action Recognition","dataset":"Charades","model":"MAXI","rank_in_archive_order":3,"of":4,"metrics":{"mAP":"23.8"},"uses_additional_data":false},{"leaderboard":"/sota/zero-shot-action-recognition-on-hmdb51","task":"Zero-Shot Action Recognition","dataset":"HMDB51","model":"MAXI","rank_in_archive_order":10,"of":29,"metrics":{"Top-1 Accuracy":"52.3"},"uses_additional_data":false},{"leaderboard":"/sota/zero-shot-action-recognition-on-kinetics","task":"Zero-Shot Action Recognition","dataset":"Kinetics","model":"MAXI","rank_in_archive_order":4,"of":20,"metrics":{"Top-1 Accuracy":"71.6"},"uses_additional_data":false},{"leaderboard":"/sota/zero-shot-action-recognition-on-ucf101","task":"Zero-Shot Action Recognition","dataset":"UCF101","model":"MAXI","rank_in_archive_order":12,"of":35,"metrics":{"Top-1 Accuracy":"78.2"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2303.08914","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2303.08914"}},"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/wlin-at/maxi","reach":{"status":"ok"}},{"provenance":"deterministic:regex_extraction","url":"https://github.com/wlin-at/MAXI","reach":{"status":"ok"}}],"summary":{"ran_draft_wrong":3,"unverified":1},"by_repo_kind":{"official":{"samples":4,"ran":3,"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":4,"samples":[{"code_sha256_prefix":"ac0675b539662e24","entry":"load_class_textbag","repo":"wlin-at/MAXI","repo_kind":"official","path":"trainers/vificlip_caption_aug_mil.py","file_url":"https://github.com/wlin-at/MAXI/blob/HEAD/trainers/vificlip_caption_aug_mil.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"ac0675b539662e24"}},{"code_sha256_prefix":"72de8381c8ab4870","entry":"load_gpt3_verbs","repo":"wlin-at/MAXI","repo_kind":"official","path":"trainers/vificlip_caption_aug_mil.py","file_url":"https://github.com/wlin-at/MAXI/blob/HEAD/trainers/vificlip_caption_aug_mil.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"72de8381c8ab4870"}},{"code_sha256_prefix":"20a487dee82777d0","entry":"load_vid_blip_textbag","repo":"wlin-at/MAXI","repo_kind":"official","path":"trainers/vificlip_caption_aug_mil.py","file_url":"https://github.com/wlin-at/MAXI/blob/HEAD/trainers/vificlip_caption_aug_mil.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"20a487dee82777d0"}},{"code_sha256_prefix":"1143df1338a3021b","entry":"ViFiCLIP_caption_aug_MIL","repo":"wlin-at/MAXI","repo_kind":"official","path":"trainers/vificlip_caption_aug_mil.py","file_url":"https://github.com/wlin-at/MAXI/blob/HEAD/trainers/vificlip_caption_aug_mil.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"1143df1338a3021b"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}