DeepFN：基于深度人脸归一化的可泛化面部动作单元识别
DeepFN: Towards Generalizable Facial Action Unit Recognition with Deep Face Normalization
AI总结：
针对面部动作单元识别对未见过的人脸和人群特征泛化能力有限的问题，提出基于自监督去噪自编码器的深度人脸归一化方法DeepFN，将表情迁移至通用模板，显著缩小了不同个体、性别、肤色和数据集间的泛化差距。
摘要：
面部动作单元识别在许多领域都有应用，从市场研究到心理治疗，从图像描述到娱乐。尽管近期取得了一些进展，但由于对未见过的人脸和人群特征的泛化能力有限，这些模型的部署受到了阻碍。本研究在多个维度上对性能进行了深入分析：个体（40名受试者）、性别（男性和女性）、肤色类型（较深和较浅）以及数据库（BP4D和DISFA）。为了帮助抑制数据中的方差，我们使用自监督去噪自编码器的概念设计了一种深度人脸归一化（DeepFN）方法，该方法将不同人的面部表情迁移到一个通用的面部模板上，随后用于训练和评估面部动作识别模型。我们表明，与特定人物模型（60.3%）相比，人物无关模型的性能显著较低（40名受试者的平均F1分数和准确率为55%），导致5.3%的泛化差距。然而，使用新引入的DeepFN对数据进行归一化，显著提高了人物无关模型的性能（59.6%），有效缩小了这一差距。同样，在考虑性别（2.4%）、肤色类型（5.3%）和数据集（9.4%）时，我们也观察到了泛化差距，而使用DeepFN显著减小了这些差距。这些发现代表了向创建更具泛化能力的面部动作单元识别系统迈出的重要一步。
英文摘要：
Facial action unit recognition has many applications from market research to psychotherapy and from image captioning to entertainment. Despite its recent progress, deployment of these models has been impeded due to their limited generalization to unseen people and demographics. This work conducts an in-depth analysis of performance across several dimensions: individuals(40 subjects), genders (male and female), skin types (darker and lighter), and databases (BP4D and DISFA). To help suppress the variance in data, we use the notion of self-supervised denoising autoencoders to design a method for deep face normalization(DeepFN) that transfers facial expressions of different people onto a common facial template which is then used to train and evaluate facial action recognition models. We show that person-independent models yield significantly lower performance (55% average F1 and accuracy across 40 subjects) than person-dependent models (60.3%), leading to a generalization gap of 5.3%. However, normalizing the data with the newly introduced DeepFN significantly increased the performance of person-independent models (59.6%), effectively reducing the gap. Similarly, we observed generalization gaps when considering gender (2.4%), skin type (5.3%), and dataset (9.4%), which were significantly reduced with the use of DeepFN. These findings represent an important step towards the creation of more generalizable facial action unit recognition systems.
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。
