Librarian Bot. I found the following papers similar to this paper.

The following papers were recommended by the Semantic Scholar API

VideoPoet: A Large Language Model for Zero-Shot Video Generation (2023)
Audio-Visual LLM for Video Understanding (2023)
Distilling Vision-Language Models on Millions of Videos (2024)
WorldDreamer: Towards General World Models for Video Generation via Predicting Masked Tokens (2024)
Latte: Latent Diffusion Transformer for Video Generation (2024)

Please give a thumbs up to this comment if you found it helpful!

If you want recommendations for any Paper on Hugging Face checkout this Space

You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: \n\n@librarian-bot\n\t recommend

\n","updatedAt":"2024-02-07T01:21:55.083Z","author":{"_id":"63d3e0e8ff1384ce6c5dd17d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1674830754237-63d3e0e8ff1384ce6c5dd17d.jpeg","fullname":"Librarian Bot (Bot)","name":"librarian-bot","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":318,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.6999517679214478},"editors":["librarian-bot"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/1674830754237-63d3e0e8ff1384ce6c5dd17d.jpeg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2402.03161","authors":[{"_id":"65c1bb8ba4c55b9d62c43b19","user":{"_id":"64b79ec0a8c39dc078897430","avatarUrl":"/avatars/8339f3ab9bb7baaf69bf174eafb7282c.svg","isPro":false,"fullname":"Yang Jin","user":"rain1011","type":"user"},"name":"Yang Jin","status":"claimed_verified","statusLastChangedAt":"2024-10-10T11:00:00.802Z","hidden":false},{"_id":"65c1bb8ba4c55b9d62c43b1a","user":{"_id":"62fc758172a7ab50b4b89c8c","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1677251053045-62fc758172a7ab50b4b89c8c.jpeg","isPro":false,"fullname":"Zhicheng Sun","user":"feifeiobama","type":"user"},"name":"Zhicheng Sun","status":"claimed_verified","statusLastChangedAt":"2024-06-25T07:34:33.836Z","hidden":false},{"_id":"65c1bb8ba4c55b9d62c43b1b","name":"Kun Xu","hidden":false},{"_id":"65c1bb8ba4c55b9d62c43b1c","name":"Kun Xu","hidden":false},{"_id":"65c1bb8ba4c55b9d62c43b1d","name":"Liwei Chen","hidden":false},{"_id":"65c1bb8ba4c55b9d62c43b1e","name":"Hao Jiang","hidden":false},{"_id":"65c1bb8ba4c55b9d62c43b1f","name":"Quzhe Huang","hidden":false},{"_id":"65c1bb8ba4c55b9d62c43b20","name":"Chengru Song","hidden":false},{"_id":"65c1bb8ba4c55b9d62c43b21","name":"Yuliang Liu","hidden":false},{"_id":"65c1bb8ba4c55b9d62c43b22","name":"Di Zhang","hidden":false},{"_id":"65c1bb8ba4c55b9d62c43b23","name":"Yang Song","hidden":false},{"_id":"65c1bb8ba4c55b9d62c43b24","name":"Kun Gai","hidden":false},{"_id":"65c1bb8ba4c55b9d62c43b25","name":"Yadong Mu","hidden":false}],"publishedAt":"2024-02-05T16:30:49.000Z","submittedOnDailyAt":"2024-02-06T02:24:39.501Z","title":"Video-LaVIT: Unified Video-Language Pre-training with Decoupled\n Visual-Motional Tokenization","submittedOnDailyBy":{"_id":"60f1abe7544c2adfd699860c","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1674929746905-60f1abe7544c2adfd699860c.jpeg","isPro":false,"fullname":"AK","user":"akhaliq","type":"user"},"summary":"In light of recent advances in multimodal Large Language Models (LLMs), there\nis increasing attention to scaling them from image-text data to more\ninformative real-world videos. Compared to static images, video poses unique\nchallenges for effective large-scale pre-training due to the modeling of its\nspatiotemporal dynamics. In this paper, we address such limitations in\nvideo-language pre-training with an efficient video decomposition that\nrepresents each video as keyframes and temporal motions. These are then adapted\nto an LLM using well-designed tokenizers that discretize visual and temporal\ninformation as a few tokens, thus enabling unified generative pre-training of\nvideos, images, and text. At inference, the generated tokens from the LLM are\ncarefully recovered to the original continuous pixel space to create various\nvideo content. Our proposed framework is both capable of comprehending and\ngenerating image and video content, as demonstrated by its competitive\nperformance across 13 multimodal benchmarks in image and video understanding\nand generation. Our code and models will be available at\nhttps://video-lavit.github.io.","upvotes":16,"discussionId":"65c1bb8fa4c55b9d62c43bc5","githubRepo":"https://github.com/jy0205/LaVIT","githubRepoAddedBy":"auto","ai_summary":"A framework for pre-training multimodal Large Language Models using video decomposition into keyframes and temporal motions achieves competitive performance across benchmarks.","ai_keywords":["multimodal Large Language Models","video decomposition","keyframes","temporal motions","tokenizers","generative pre-training","video content","image and video understanding","image and video generation"],"githubStars":603},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"632656382858ff914befa0b6","avatarUrl":"/avatars/979772338dd29c722e6ad92620c2e65c.svg","isPro":false,"fullname":"Kel","user":"Pepperino","type":"user"},{"_id":"61868ce808aae0b5499a2a95","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/61868ce808aae0b5499a2a95/F6BA0anbsoY_Z7M1JrwOe.jpeg","isPro":true,"fullname":"Sylvain Filoni","user":"fffiloni","type":"user"},{"_id":"6335604ea01bd734f72316b0","avatarUrl":"/avatars/4c6611dabd492106ffb2e82fd680d983.svg","isPro":false,"fullname":"Zhizhou Sha","user":"JamesSand","type":"user"},{"_id":"6032802e1f993496bc14d9e3","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6032802e1f993496bc14d9e3/w6hr-DEQot4VVkoyRIBiy.png","isPro":false,"fullname":"Omar Sanseviero","user":"osanseviero","type":"user"},{"_id":"6079cc1c65b9d0165cb18394","avatarUrl":"/avatars/8c1f1011d9f675fc899919cf07faef68.svg","isPro":false,"fullname":"Chris Lesniewski","user":"lesniewski","type":"user"},{"_id":"62fc758172a7ab50b4b89c8c","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1677251053045-62fc758172a7ab50b4b89c8c.jpeg","isPro":false,"fullname":"Zhicheng Sun","user":"feifeiobama","type":"user"},{"_id":"620783f24e28382272337ba4","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/620783f24e28382272337ba4/zkUveQPNiDfYjgGhuFErj.jpeg","isPro":false,"fullname":"GuoLiangTang","user":"Tommy930","type":"user"},{"_id":"6538119803519fddb4a17e10","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6538119803519fddb4a17e10/ffJMkdx-rM7VvLTCM6ri_.jpeg","isPro":false,"fullname":"samusenps","user":"samusenps","type":"user"},{"_id":"6300afbba123c93a5fabefeb","avatarUrl":"/avatars/fca3e62fe82a15f88529d5843dd17373.svg","isPro":false,"fullname":"Songtao liu","user":"ruinmessi","type":"user"},{"_id":"63913b120cf6b11c487ca31d","avatarUrl":"/avatars/aec44edd5470dd6e767e0a25efd6fb5d.svg","isPro":true,"fullname":"Xin Li","user":"lixin4ever","type":"user"},{"_id":"635cada2c017767a629db012","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1667018139063-noauth.jpeg","isPro":false,"fullname":"Ojasvi Singh Yadav","user":"ojasvisingh786","type":"user"},{"_id":"635964636a61954080850e1d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/635964636a61954080850e1d/0bfExuDTrHTtm8c-40cDM.png","isPro":false,"fullname":"William Lamkin","user":"phanes","type":"user"}],"acceptLanguages":["*"],"dailyPaperRank":0}">

Papers

arxiv:2402.03161

Video-LaVIT: Unified Video-Language Pre-training with Decoupled Visual-Motional Tokenization

Published on Feb 5, 2024

· Submitted by

AK on Feb 6, 2024

Upvote

Authors:

Yang Jin ,

Zhicheng Sun ,

Abstract

A framework for pre-training multimodal Large Language Models using video decomposition into keyframes and temporal motions achieves competitive performance across benchmarks.

AI-generated summary

In light of recent advances in multimodal Large Language Models (LLMs), there is increasing attention to scaling them from image-text data to more informative real-world videos. Compared to static images, video poses unique challenges for effective large-scale pre-training due to the modeling of its spatiotemporal dynamics. In this paper, we address such limitations in video-language pre-training with an efficient video decomposition that represents each video as keyframes and temporal motions. These are then adapted to an LLM using well-designed tokenizers that discretize visual and temporal information as a few tokens, thus enabling unified generative pre-training of videos, images, and text. At inference, the generated tokens from the LLM are carefully recovered to the original continuous pixel space to create various video content. Our proposed framework is both capable of comprehending and generating image and video content, as demonstrated by its competitive performance across 13 multimodal benchmarks in image and video understanding and generation. Our code and models will be available at https://video-lavit.github.io.