arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2303.01776cs.CV
收藏

基于先验信息的分解重构学习微表情识别

Prior Information based Decomposition and Reconstruction Learning for Micro-Expression Recognition

Nanjing University of Posts and Telecommunications(南京邮电大学);Nanjing Institute of Technology(南京工程学院)

Jinsheng Wei, Haoyu Chen, Guanming Lu, Jingjie Yan, Yue Xie, Guoying Zhao

摘要:

微表情识别(MER)吸引了广泛的研究兴趣,因为微表情(ME)可以推断真实的情感。先验信息可以指导模型有效地学习判别性ME特征。然而,大多数工作集中在研究具有更强的表示能力的一般模型,以适应性地聚合ME运动信息的整体方式,这可能会忽略的先验信息和属性的ME。为了解决这个问题,在先验信息的驱动下,ME的类别可以通过面部不同组件之间的动作之间的关系来推断,这项工作设计了一种新的模型,可以符合这种先验信息,并学习ME运动特征,以可解释的方式。具体而言,本文提出了一种基于分解和重构的图表示学习(DeRe-GRL)模型,以有效地学习高级别ME功能。DeRe-GRL包括两个模块:动作分解模块(ADM)和关系重构模块(RRM),其中ADM学习面部关键部件的动作特征,RRM探索这些动作特征之间的关系。ADM基于人脸关键组件,将基于图模型的骨干提取的几何运动特征划分为若干个子特征,学习映射矩阵将这些子特征映射为多个动作特征; RRM学习权值对所有动作特征进行加权,建立动作特征之间的关系。实验结果证明了所提出的模块的有效性,所提出的方法取得了竞争力的性能。

英文摘要:

Micro-expression recognition (MER) draws intensive research interest as micro-expressions (MEs) can infer genuine emotions. Prior information can guide the model to learn discriminative ME features effectively. However, most works focus on researching the general models with a stronger representation ability to adaptively aggregate ME movement information in a holistic way, which may ignore the prior information and properties of MEs. To solve this issue, driven by the prior information that the category of ME can be inferred by the relationship between the actions of facial different components, this work designs a novel model that can conform to this prior information and learn ME movement features in an interpretable way. Specifically, this paper proposes a Decomposition and Reconstruction-based Graph Representation Learning (DeRe-GRL) model to effectively learn high-level ME features. DeRe-GRL includes two modules: Action Decomposition Module (ADM) and Relation Reconstruction Module (RRM), where ADM learns action features of facial key components and RRM explores the relationship between these action features. Based on facial key components, ADM divides the geometric movement features extracted by the graph model-based backbone into several sub-features, and learns the map matrix to map these sub-features into multiple action features; then, RRM learns weights to weight all action features to build the relationship between action features. The experimental results demonstrate the effectiveness of the proposed modules, and the proposed method achieves competitive performance.

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。

↑