{"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/video-llava-learning-united-visual-1","title":"Video-LLaVA: Learning United Visual Representation by Alignment Before Projection","arxiv_id":"2311.10122","date":"2023-11-16","proceeding":null,"authors":["Bin Lin","Yang Ye","Bin Zhu","Jiaxi Cui","Munan Ning","Peng Jin","Li Yuan"],"abstract":"The Large Vision-Language Model (LVLM) has enhanced the performance of various downstream tasks in visual-language understanding. Most existing approaches encode images and videos into separate feature spaces, which are then fed as inputs to large language models. However, due to the lack of unified tokenization for images and videos, namely misalignment before projection, it becomes challenging for a Large Language Model (LLM) to learn multi-modal interactions from several poor projection layers. In this work, we unify visual representation into the language feature space to advance the foundational LLM towards a unified LVLM. As a result, we establish a simple but robust LVLM baseline, Video-LLaVA, which learns from a mixed dataset of images and videos, mutually enhancing each other. Video-LLaVA achieves superior performances on a broad range of 9 image benchmarks across 5 image question-answering datasets and 4 image benchmark toolkits. Additionally, our Video-LLaVA also outperforms Video-ChatGPT by 5.8%, 9.9%, 18.6%, and 10.1% on MSRVTT, MSVD, TGIF, and ActivityNet, respectively. Notably, extensive experiments demonstrate that Video-LLaVA mutually benefits images and videos within a unified visual representation, outperforming models designed specifically for images or videos. We aim for this work to provide modest insights into the multi-modal inputs for the LLM. Code address: \\href{https://github.com/PKU-YuanGroup/Video-LLaVA}","url_abs":"https://arxiv.org/abs/2311.10122v3","url_pdf":"https://arxiv.org/pdf/2311.10122v3.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":"video-llava-learning-united-visual-1","repo_url":"https://github.com/PKU-YuanGroup/Video-LLaVA","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"video-llava-learning-united-visual-1","repo_url":"https://github.com/PKU-YuanGroup/LLMBind","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"video-llava-learning-united-visual-1","repo_url":"https://github.com/PKU-YuanGroup/MoE-LLaVA","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"video-llava-learning-united-visual-1","repo_url":"https://github.com/pku-yuangroup/languagebind","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"video-llava-learning-united-visual-1","repo_url":"https://github.com/pku-yuangroup/video-bench","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}},{"paper_slug":"video-llava-learning-united-visual-1","repo_url":"https://github.com/qiujihao19/artemis","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"language-modeling","task_name":"Language Modeling"},{"task_slug":"language-modelling","task_name":"Language Modelling"},{"task_slug":"large-language-model","task_name":"Large Language Model"},{"task_slug":"multiple-choice","task_name":"Multiple-choice"},{"task_slug":"question-answering","task_name":"Question Answering"},{"task_slug":"temporal-relation-extraction","task_name":"Temporal Relation Extraction"},{"task_slug":"video-question-answering","task_name":"Video Question Answering"},{"task_slug":"visual-question-answering-1","task_name":"Visual Question Answering"},{"task_slug":"visual-question-answering","task_name":"Visual Question Answering (VQA)"},{"task_slug":"zeroshot-video-question-answer","task_name":"Zero-Shot Video Question Answer"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/temporal-relation-extraction-on-vinoground","task":"Temporal Relation Extraction","dataset":"Vinoground","model":"Video-LLaVA-7B","rank_in_archive_order":15,"of":24,"metrics":{"Group Score":"6.6","Text Score":"24.8","Video Score":"25.8"},"uses_additional_data":false},{"leaderboard":"/sota/video-question-answering-on-activitynet-qa","task":"Video Question Answering","dataset":"ActivityNet-QA","model":"Video-LLaVA","rank_in_archive_order":16,"of":36,"metrics":{"Accuracy":"45.3","Confidence score":"3.3"},"uses_additional_data":false},{"leaderboard":"/sota/visual-question-answering-on-mm-vet","task":"Visual Question Answering","dataset":"MM-Vet","model":"Video-LLaVA","rank_in_archive_order":187,"of":231,"metrics":{"GPT-4 score":"32.0"},"uses_additional_data":false},{"leaderboard":"/sota/zeroshot-video-question-answer-on-activitynet","task":"Zero-Shot Video Question Answer","dataset":"ActivityNet-QA","model":"Video-LLaVA","rank_in_archive_order":22,"of":28,"metrics":{"Accuracy":"45.3","Confidence Score":"3.3"},"uses_additional_data":false},{"leaderboard":"/sota/zeroshot-video-question-answer-on-msrvtt-qa","task":"Zero-Shot Video Question Answer","dataset":"MSRVTT-QA","model":"Video-LLaVA-7B","rank_in_archive_order":17,"of":30,"metrics":{"Accuracy":"59.2","Confidence Score":"3.5"},"uses_additional_data":true},{"leaderboard":"/sota/zeroshot-video-question-answer-on-msvd-qa","task":"Zero-Shot Video Question Answer","dataset":"MSVD-QA","model":"Video-LLaVA-7B","rank_in_archive_order":17,"of":28,"metrics":{"Accuracy":"70.7","Confidence Score":"3.9"},"uses_additional_data":false},{"leaderboard":"/sota/zeroshot-video-question-answer-on-tgif-qa","task":"Zero-Shot Video Question Answer","dataset":"TGIF-QA","model":"Video-LLaVA-7B","rank_in_archive_order":9,"of":14,"metrics":{"Accuracy":"70.0","Confidence Score":"4.0"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2311.10122","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2311.10122"}},"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/PKU-YuanGroup/Video-LLaVA","reach":{"status":"ok","spdx":"Apache-2.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/pku-yuangroup/video-bench","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/PKU-YuanGroup/MoE-LLaVA","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/qiujihao19/artemis","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/pku-yuangroup/languagebind","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/PKU-YuanGroup/LLMBind","reach":null}],"summary":{"ran":3,"ran_honours":1,"unverified":3},"by_repo_kind":{"listed":{"samples":6,"ran":4,"repositories":2}},"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":1,"samples":[{"code_sha256_prefix":"ac6a026c5d4ff47b","entry":"get_depth_transform","repo":"pku-yuangroup/languagebind","repo_kind":"listed","path":"languagebind/depth/processing_depth.py","file_url":"https://github.com/pku-yuangroup/languagebind/blob/HEAD/languagebind/depth/processing_depth.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"ac6a026c5d4ff47b"}},{"code_sha256_prefix":"20a58a8c04ccda43","entry":"int16_to_float32_torch","repo":"pku-yuangroup/languagebind","repo_kind":"listed","path":"languagebind/audio/processing_audio.py","file_url":"https://github.com/pku-yuangroup/languagebind/blob/HEAD/languagebind/audio/processing_audio.py","link_basis":"harvester_set","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"20a58a8c04ccda43"}},{"code_sha256_prefix":"5530423121ac4757","entry":"opencv_loader","repo":"pku-yuangroup/languagebind","repo_kind":"listed","path":"languagebind/depth/processing_depth.py","file_url":"https://github.com/pku-yuangroup/languagebind/blob/HEAD/languagebind/depth/processing_depth.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"5530423121ac4757"}},{"code_sha256_prefix":"833effea5806a36e","entry":"stack_dict","repo":"pku-yuangroup/languagebind","repo_kind":"listed","path":"model/languagebind.py","file_url":"https://github.com/pku-yuangroup/languagebind/blob/HEAD/model/languagebind.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"833effea5806a36e"}},{"code_sha256_prefix":"3d9df14425d0ebcf","entry":"LlavaMetaForCausalLM","repo":"PKU-YuanGroup/LLMBind","repo_kind":"listed","path":"model/llava/model/llava_arch.py","file_url":"https://github.com/PKU-YuanGroup/LLMBind/blob/HEAD/model/llava/model/llava_arch.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"3d9df14425d0ebcf"}},{"code_sha256_prefix":"6d6b281af73773f5","entry":"build_video_tower","repo":null,"repo_kind":null,"path":null,"file_url":null,"link_basis":"identical_code_first_harvested_elsewhere","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":null,"inline_ok":false,"mcp_get_code":{"code_sha256":"6d6b281af73773f5"}},{"code_sha256_prefix":"852e87ab903330a7","entry":"make_list_of_images","repo":"pku-yuangroup/languagebind","repo_kind":"listed","path":"languagebind/audio/processing_audio.py","file_url":"https://github.com/pku-yuangroup/languagebind/blob/HEAD/languagebind/audio/processing_audio.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":"852e87ab903330a7"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}