arXivDaily arXiv每日学术速递 周一至周五更新
北美退伍军人节(Veterans Day),12 日无更新
arXiv 2002.06353cs.CVcs.CLcs.LGeess.ASeess.IV
收藏

UniVL:一个统一的多模态理解和生成的视频和语言预训练模型

UniVL: A Unified Video and Language Pre-Training Model for Multimodal Understanding and Generation

Southwest Jiaotong University(西南交通大学);Microsoft Research Asia(微软亚洲研究院);Institute of Computing Technology, Chinese Academy of Science(中国科学院计算技术研究所);University of Chinese Academy of Sciences(中国科学院大学);Beijing Institute of Technology(北京理工大学);Microsoft STCA(微软(中国)有限公司)

Huaishao Luo, Lei Ji, Botian Shi, Haoyang Huang, Nan Duan, Tianrui Li, Jason Li, Taroon Bharti, Ming Zhou

摘要:

随着NLP和图像语言任务预训练技术的成功,一些视频语言预训练工作逐渐发展起来,以改善视频文本相关的下游任务。然而,大多数现有的多模态模型是预先训练的理解任务,导致生成任务的预训练微调差异。本文提出了UniVL:一个统一的视频和语言预训练模型,用于多模态理解和生成。它由四个组件组成,包括两个单模编码器,一个交叉编码器和一个具有Transformer主干的解码器。设计了五个目标,包括视频-文本连接,条件掩蔽语言模型(CMLM),条件掩蔽框架模型(CMFM),视频-文本对齐和语言重建,训练每个组件。我们进一步开发了两种预训练策略,逐步预训练(StagedP)和增强的视频表示(EnhancedV),以使UniVL的训练过程更有效。预训练在一个相当大的教学视频数据集HowTo 100 M上进行。实验结果表明,UniVL可以学习强大的视频-文本表示,并在五个下游任务上取得了最先进的结果。

英文摘要:

With the recent success of the pre-training technique for NLP and image-linguistic tasks, some video-linguistic pre-training works are gradually developed to improve video-text related downstream tasks. However, most of the existing multimodal models are pre-trained for understanding tasks, leading to a pretrain-finetune discrepancy for generation tasks. This paper proposes UniVL: a Unified Video and Language pre-training model for both multimodal understanding and generation. It comprises four components, including two single-modal encoders, a cross encoder, and a decoder with the Transformer backbone. Five objectives, including video-text joint, conditioned masked language model (CMLM), conditioned masked frame model (CMFM), video-text alignment, and language reconstruction, are designed to train each of the components. We further develop two pre-training strategies, stage by stage pre-training (StagedP) and enhanced video representation (EnhancedV), to make the training process of the UniVL more effective. The pre-train is carried out on a sizeable instructional video dataset HowTo100M. Experimental results demonstrate that the UniVL can learn strong video-text representation and achieves state-of-the-art results on five downstream tasks.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑