arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2004.06657cs.CV
收藏

基于迁移学习的热图回归动作单元强度估计

A Transfer Learning approach to Heatmap Regression for Action Unit intensity estimation

University of Nottingham(诺丁汉大学);Queen Mary University of London(伦敦玛丽女王大学);Samsung AI Center Cambridge(三星剑桥人工智能中心)

Ioanna Ntinou, Enrique Sanchez, Adrian Bulat, Michel Valstar, Georgios Tzimiropoulos

摘要:

动作单元(AU)是已知在特定面部位置产生外观变化的基于几何学的原子面部肌肉运动。出于这种观察,我们提出了一种新的AU建模问题,包括共同估计其定位和强度。为此,我们提出了一种基于热图回归的简单而有效的方法,将这两个问题合并到一个任务中。热图对AU是否出现在给定空间位置处进行建模。为了适应AU强度的联合建模,我们提出了可变大小的热图,其幅度和大小根据标记的强度而变化。使用热图回归,我们可以继承最近在面部标志定位方面所取得的进展。基于这两个问题之间的相似性,我们设计了一种迁移学习方法,利用在大规模面部标志数据集上训练的网络知识。特别是,我们通过a)微调,b)适应层,c)注意力地图和d)重新参数化来探索迁移学习的不同替代方案。我们的方法有效地继承了强大的面部对齐网络产生的丰富的面部特征,以最小的额外计算成本。我们通过经验验证,我们的系统在三个流行的数据集上设置了新的最先进的技术,即BP4D,DISFA和FERA2017。

英文摘要:

Action Units (AUs) are geometrically-based atomic facial muscle movements known to produce appearance changes at specific facial locations. Motivated by this observation we propose a novel AU modelling problem that consists of jointly estimating their localisation and intensity. To this end, we propose a simple yet efficient approach based on Heatmap Regression that merges both problems into a single task. A Heatmap models whether an AU occurs or not at a given spatial location. To accommodate the joint modelling of AUs intensity, we propose variable size heatmaps, with their amplitude and size varying according to the labelled intensity. Using Heatmap Regression, we can inherit from the progress recently witnessed in facial landmark localisation. Building upon the similarities between both problems, we devise a transfer learning approach where we exploit the knowledge of a network trained on large-scale facial landmark datasets. In particular, we explore different alternatives for transfer learning through a) fine-tuning, b) adaptation layers, c) attention maps, and d) reparametrisation. Our approach effectively inherits the rich facial features produced by a strong face alignment network, with minimal extra computational cost. We empirically validate that our system sets a new state-of-the-art on three popular datasets, namely BP4D, DISFA, and FERA2017.

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。

↑