Timezone: »

 
Poster
Expectation-Maximization Contrastive Learning for Compact Video-and-Language Representations
Peng Jin · Jinfa Huang · Fenglin Liu · Xian Wu · Shen Ge · Guoli Song · David Clifton · Jie Chen

@

Most video-and-language representation learning approaches employ contrastive learning, e.g., CLIP, to project the video and text features into a common latent space according to the semantic similarities of text-video pairs. However, such learned shared latent spaces are not often optimal, and the modality gap between visual and textual representation can not be fully eliminated. In this paper, we propose Expectation-Maximization Contrastive Learning (EMCL) to learn compact video-and-language representations. Specifically, we use the Expectation-Maximization algorithm to find a compact set of bases for the latent space, where the features could be concisely represented as the linear combinations of these bases. Such feature decomposition of video-and-language representations reduces the rank of the latent space, resulting in increased representing power for the semantics. Extensive experiments on three benchmark text-video retrieval datasets prove that our EMCL can learn more discriminative video-and-language representations than previous methods, and significantly outperform previous state-of-the-art methods across all metrics. More encouragingly, the proposed method can be applied to boost the performance of existing approaches either as a jointly training layer or an out-of-the-box inference module with no extra training, making it easy to be incorporated into any existing methods.

Keywords: contrastive learning · expectation-maximization algorithm · cross-modal retrieval · video-and-language representation learning
[ Paper]  [ Poster]  [ OpenReview] 

Author Information

Peng Jin (Peking University)
Jinfa Huang (Peking University)
Fenglin Liu (University of Oxford)
Xian Wu (Tencent)
Shen Ge (Tencent)
Guoli Song (Peng Cheng Laboratory)
David Clifton (University of Oxford)

Professor of Clinical Machine Learning Department of Engineering Science University of Oxford

Jie Chen (Peng Cheng Laboratory)

More from the Same Authors