MemBridge: Video-Language Pre-Training With Memory-Augmented Inter-Modality Bridge

doi:10.1109/TIP.2023.3283916

	MemBridge: Video-Language Pre-Training With Memory-Augmented Inter-Modality Bridge
	Yang, Jiahao 1,2; Li, Xiangyang 1,2; Zheng, Mao 3; Wang, Zihan 1,2; Zhu, Yongqing 1,2; Guo, Xiaoqian 1,2; Yuan, Yuchen 3; Chai, Zifeng 3; Jiang, Shuqiang 1,2
	2023
发表期刊	IEEE TRANSACTIONS ON IMAGE PROCESSING
ISSN	1057-7149
卷号	32 页码:4073-4087
摘要	Video-language pre-training has attracted considerable attention recently for its promising performance on various downstream tasks. Most existing methods utilize the modality-specific or modality-joint representation architectures for the cross-modality pre-training. Different from previous methods, this paper presents a novel architecture named Memory-augmented Inter-Modality Bridge (MemBridge), which uses the learnable intermediate modality representations as the bridge for the interaction between videos and language. Specifically, in the transformer-based cross-modality encoder, we introduce the learnable bridge tokens as the interaction approach, which means the video and language tokens can only perceive information from bridge tokens and themselves. Moreover, a memory bank is proposed to store abundant modality interaction information for adaptively generating bridge tokens according to different cases, enhancing the capacity and robustness of the inter-modality bridge. Through pre-training, MemBridge explicitly models the representations for more sufficient inter-modality interaction. Comprehensive experiments show that our approach achieves competitive performance with previous methods on various downstream tasks including video-text retrieval, video captioning, and video question answering on multiple datasets, demonstrating the effectiveness of the proposed method. The code has been available at https://github.com/jahhaoyang/MemBridge.
关键词	Video-language pre-training inter-modality bridge memory module
DOI	10.1109/TIP.2023.3283916
收录类别	SCI
语种	英语
WOS研究方向	Computer Science ; Engineering
WOS类目	Computer Science, Artificial Intelligence ; Engineering, Electrical & Electronic
WOS记录号	WOS:001033515600013
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
引用统计
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/21342
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Jiang, Shuqiang
作者单位	1.Chinese Acad Sci, Inst Comp Technol, Key Lab Intelligent Informat Proc, Beijing 100190, Peoples R China 2.Univ Chinese Acad Sci, Beijing 100049, Peoples R China 3.Tencent Co Ltd, Dept Machine Learning Platform, Beijing 100193, Peoples R China
推荐引用方式 GB/T 7714	Yang, Jiahao,Li, Xiangyang,Zheng, Mao,et al. MemBridge: Video-Language Pre-Training With Memory-Augmented Inter-Modality Bridge[J]. IEEE TRANSACTIONS ON IMAGE PROCESSING,2023,32:4073-4087.
APA	Yang, Jiahao.,Li, Xiangyang.,Zheng, Mao.,Wang, Zihan.,Zhu, Yongqing.,...&Jiang, Shuqiang.(2023).MemBridge: Video-Language Pre-Training With Memory-Augmented Inter-Modality Bridge.IEEE TRANSACTIONS ON IMAGE PROCESSING,32,4073-4087.
MLA	Yang, Jiahao,et al."MemBridge: Video-Language Pre-Training With Memory-Augmented Inter-Modality Bridge".IEEE TRANSACTIONS ON IMAGE PROCESSING 32(2023):4073-4087.