Efficiente Tokenizzazione di Video Lunghi tramite Ricostruzione di Patch basata su Coordinate.
Efficient Long Video Tokenization via Coordinated-based Patch Reconstruction
November 22, 2024
Autori: Huiwon Jang, Sihyun Yu, Jinwoo Shin, Pieter Abbeel, Younggyo Seo
cs.AI
Abstract
L'efficiente tokenizzazione dei video rimane una sfida nel training dei modelli di visione in grado di processare video lunghi. Una direzione promettente è lo sviluppo di un tokenizer in grado di codificare lunghi spezzoni video, poiché consentirebbe al tokenizer di sfruttare meglio la coerenza temporale dei video per la tokenizzazione. Tuttavia, addestrare i tokenizzatori esistenti su video lunghi comporta spesso un enorme costo di addestramento poiché vengono addestrati a ricostruire tutti i frame in una volta. In questo articolo, presentiamo CoordTok, un video tokenizer che apprende un mapping da rappresentazioni basate sulle coordinate alle patch corrispondenti dei video di input, ispirato ai recenti progressi nei modelli generativi 3D. In particolare, CoordTok codifica un video in rappresentazioni triplane fattorizzate e ricostruisce patch corrispondenti a coordinate (x,y,t) campionate casualmente. Ciò consente di addestrare grandi modelli di tokenizer direttamente su video lunghi senza richiedere risorse di addestramento eccessive. I nostri esperimenti mostrano che CoordTok può ridurre drasticamente il numero di token per codificare lunghi spezzoni video. Ad esempio, CoordTok può codificare un video di 128 frame con risoluzione 128x128 in 1280 token, mentre i baselines necessitano di 6144 o 8192 token per ottenere una qualità di ricostruzione simile. Mostriamo inoltre che questa efficiente tokenizzazione dei video consente un addestramento efficiente in termini di memoria di un transformer a diffusione in grado di generare 128 frame in una sola volta.
English
Efficient tokenization of videos remains a challenge in training vision
models that can process long videos. One promising direction is to develop a
tokenizer that can encode long video clips, as it would enable the tokenizer to
leverage the temporal coherence of videos better for tokenization. However,
training existing tokenizers on long videos often incurs a huge training cost
as they are trained to reconstruct all the frames at once. In this paper, we
introduce CoordTok, a video tokenizer that learns a mapping from
coordinate-based representations to the corresponding patches of input videos,
inspired by recent advances in 3D generative models. In particular, CoordTok
encodes a video into factorized triplane representations and reconstructs
patches that correspond to randomly sampled (x,y,t) coordinates. This allows
for training large tokenizer models directly on long videos without requiring
excessive training resources. Our experiments show that CoordTok can
drastically reduce the number of tokens for encoding long video clips. For
instance, CoordTok can encode a 128-frame video with 128times128 resolution
into 1280 tokens, while baselines need 6144 or 8192 tokens to achieve similar
reconstruction quality. We further show that this efficient video tokenization
enables memory-efficient training of a diffusion transformer that can generate
128 frames at once.Summary
AI-Generated Summary
DeepSeek-R1: Incentivizzare la capacità di ragionamento nei LLM tramite Apprendimento per RinforzoDeepSeek-R1: Incentivizing Reasoning Capability in LLMs via
Reinforcement Learning
DeepSeek-R1: Incentivizzare la capacità di ragionamento nei LLM tramite Apprendimento per Rinforzo
DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via
Reinforcement Learning
DeepSeek-AI, Daya Guo, Dejian Yang, Haowei Zhang, Junxiao Song, Ruoyu Zhang, Runxin Xu, Qihao Zhu, Shirong Ma, Peiyi Wang, Xiao Bi, Xiaokang Zhang, Xingkai Yu, Yu Wu, Z. F. Wu, Zhibin Gou, Zhihong Shao, Zhuoshu Li, Ziyi Gao, Aixin Liu, Bing Xue, Bingxuan Wang, Bochao Wu, Bei Feng, Chengda Lu, Chenggang Zhao, Chengqi Deng, Chenyu Zhang, Chong Ruan, Damai Dai, Deli Chen, Dongjie Ji, Erhang Li, Fangyun Lin, Fucong Dai, Fuli Luo, Guangbo Hao, Guanting Chen, Guowei Li, H. Zhang, Han Bao, Hanwei Xu, Haocheng Wang, Honghui Ding, Huajian Xin, Huazuo Gao, Hui Qu, Hui Li, Jianzhong Guo, Jiashi Li, Jiawei Wang, Jingchang Chen, Jingyang Yuan, Junjie Qiu, Junlong Li, J. L. Cai, Jiaqi Ni, Jian Liang, Jin Chen, Kai Dong, Kai Hu, Kaige Gao, Kang Guan, Kexin Huang, Kuai Yu, Lean Wang, Lecong Zhang, Liang Zhao, Litong Wang, Liyue Zhang, Lei Xu, Leyi Xia, Mingchuan Zhang, Minghua Zhang, Minghui Tang, Meng Li, Miaojun Wang, Mingming Li, Ning Tian, Panpan Huang, Peng Zhang, Qiancheng Wang, Qinyu Chen, Qiushi Du, Ruiqi Ge, Ruisong Zhang, Ruizhe Pan, Runji Wang, R. J. Chen, R. L. Jin, Ruyi Chen, Shanghao Lu, Shangyan Zhou, Shanhuang Chen, Shengfeng Ye, Shiyu Wang, Shuiping Yu, Shunfeng Zhou, Shuting Pan, S. S. Li, Shuang Zhou, Shaoqing Wu, Shengfeng Ye, Tao Yun, Tian Pei, Tianyu Sun, T. Wang, Wangding Zeng, Wanjia Zhao, Wen Liu, Wenfeng Liang, Wenjun Gao, Wenqin Yu, Wentao Zhang, W. L. Xiao, Wei An, Xiaodong Liu, Xiaohan Wang, Xiaokang Chen, Xiaotao Nie, Xin Cheng, Xin Liu, Xin Xie, Xingchao Liu, Xinyu Yang, Xinyuan Li, Xuecheng Su, Xuheng Lin, X. Q. Li, Xiangyue Jin, Xiaojin Shen, Xiaosha Chen, Xiaowen Sun, Xiaoxiang Wang, Xinnan Song, Xinyi Zhou, Xianzu Wang, Xinxia Shan, Y. K. Li, Y. Q. Wang, Y. X. Wei, Yang Zhang, Yanhong Xu, Yao Li, Yao Zhao, Yaofeng Sun, Yaohui Wang, Yi Yu, Yichao Zhang, Yifan Shi, Yiliang Xiong, Ying He, Yishi Piao, Yisong Wang, Yixuan Tan, Yiyang Ma, Yiyuan Liu, Yongqiang Guo, Yuan Ou, Yuduan Wang, Yue Gong, Yuheng Zou, Yujia He, Yunfan Xiong, Yuxiang Luo, Yuxiang You, Yuxuan Liu, Yuyang Zhou, Y. X. Zhu, Yanhong Xu, Yanping Huang, Yaohui Li, Yi Zheng, Yuchen Zhu, Yunxian Ma, Ying Tang, Yukun Zha, Yuting Yan, Z. Z. Ren, Zehui Ren, Zhangli Sha, Zhe Fu, Zhean Xu, Zhenda Xie, Zhengyan Zhang, Zhewen Hao, Zhicheng Ma, Zhigang Yan, Zhiyu Wu, Zihui Gu, Zijia Zhu, Zijun Liu, Zilin Li, Ziwei Xie, Ziyang Song, Zizheng Pan, Zhen Huang, Zhipeng Xu, Zhongyu Zhang, Zhen Zhang•Jan 22, 2025•3685
Rapporto Tecnico Qwen2.5Qwen2.5 Technical Report
Rapporto Tecnico Qwen2.5
Qwen2.5 Technical Report
Qwen, An Yang, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chengyuan Li, Dayiheng Liu, Fei Huang, Haoran Wei, Huan Lin, Jian Yang, Jianhong Tu, Jianwei Zhang, Jianxin Yang, Jiaxi Yang, Jingren Zhou, Junyang Lin, Kai Dang, Keming Lu, Keqin Bao, Kexin Yang, Le Yu, Mei Li, Mingfeng Xue, Pei Zhang, Qin Zhu, Rui Men, Runji Lin, Tianhao Li, Tingyu Xia, Xingzhang Ren, Xuancheng Ren, Yang Fan, Yang Su, Yichang Zhang, Yu Wan, Yuqiong Liu, Zeyu Cui, Zhenru Zhang, Zihan Qiu•Dec 19, 2024•36311
MiniMax-01: Scalare i modelli di base con attenzione lampeggianteMiniMax-01: Scaling Foundation Models with Lightning Attention
MiniMax-01: Scalare i modelli di base con attenzione lampeggiante
MiniMax-01: Scaling Foundation Models with Lightning Attention
MiniMax, Aonian Li, Bangwei Gong, Bo Yang, Boji Shan, Chang Liu, Cheng Zhu, Chunhao Zhang, Congchao Guo, Da Chen, Dong Li, Enwei Jiao, Gengxin Li, Guojun Zhang, Haohai Sun, Houze Dong, Jiadai Zhu, Jiaqi Zhuang, Jiayuan Song, Jin Zhu, Jingtao Han, Jingyang Li, Junbin Xie, Junhao Xu, Junjie Yan, Kaishun Zhang, Kecheng Xiao, Kexi Kang, Le Han, Leyang Wang, Lianfei Yu, Liheng Feng, Lin Zheng, Linbo Chai, Long Xing, Meizhi Ju, Mingyuan Chi, Mozhi Zhang, Peikai Huang, Pengcheng Niu, Pengfei Li, Pengyu Zhao, Qi Yang, Qidi Xu, Qiexiang Wang, Qin Wang, Qiuhui Li, Ruitao Leng, Shengmin Shi, Shuqi Yu, Sichen Li, Songquan Zhu, Tao Huang, Tianrun Liang, Weigao Sun, Weixuan Sun, Weiyu Cheng, Wenkai Li, Xiangjun Song, Xiao Su, Xiaodong Han, Xinjie Zhang, Xinzhu Hou, Xu Min, Xun Zou, Xuyang Shen, Yan Gong, Yingjie Zhu, Yipeng Zhou, Yiran Zhong, Yongyi Hu, Yuanxiang Fan, Yue Yu, Yufeng Yang, Yuhao Li, Yunan Huang, Yunji Li, Yunpeng Huang, Yunzhi Xu, Yuxin Mao, Zehan Li, Zekang Li, Zewei Tao, Zewen Ying, Zhaoyang Cong, Zhen Qin, Zhenhua Fan, Zhihang Yu, Zhuo Jiang, Zijia Wu•Jan 14, 2025•2836