arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2607.04490cs.CV
收藏

利用合成数据增强头戴式显示器中的面部表情识别

Enhancing Facial Expression Recognition in Head-Mounted Displays with Synthetic Data

University of Pittsburgh(匹兹堡大学)

Jianing Deng, Qiang Zhou, Jingtong Hu

AI总结:

针对头戴式显示器中面部表情识别数据稀缺问题,提出数据合成框架,从正视图图像生成头戴式相机视图图像,经3D重建、相机系统渲染及纹理空间对齐网络,实验表明合成数据训练模型效果更好。

摘要:

面部表情识别(FER)在使用头戴式显示器(HMD)的混合现实环境中对社交互动至关重要。然而,由于隐私问题和HMD平台的多样性,从头戴式相机(HMC)收集FER数据具有挑战性。此外,由于HMC的独特视角,现有的FER数据集无法直接应用。数据稀缺阻碍了基于神经网络的HMC FER方法的发展。为了解决数据稀缺问题,我们提出了一个数据合成框架,该框架利用丰富的现有注释数据集从正视图图像生成HMC视图图像。具体来说,我们首先从图像重建3D纹理网格,然后应用可配置的相机系统从HMC视角渲染图像。此外,我们引入了一个纹理空间对齐网络(TSAN),该网络能够从图像中进行精确的纹理采样,以保留详细的面部表情。为了评估所提出的方法,我们在模拟和真实的HMC数据集上进行了广泛的实验。实验结果表明,在我们的合成数据集上训练的模型优于在现有数据集上训练的模型,并且在不同的相机配置下表现出更好的泛化能力。

英文摘要:

Facial expression recognition (FER) is crucial for social interaction in mixed reality environments that employ head-mounted displays (HMD). However, collecting FER data from head-mounted cameras (HMC) is challenging due to privacy concerns and the diversity of HMD platforms. Moreover, existing FER datasets are not directly applicable due to the unique perspectives of HMCs. The lack of sufficient data hinders the development of neural network-based HMC FER methods. To address data scarcity, we propose a data synthesis framework that generates HMC-view images from frontal-view images, leveraging abundant existing annotated datasets. Specifically, we first reconstruct 3D textured meshes from images and then apply a configurable camera system to render images from the HMC perspective. Additionally, we introduce a texture-space alignment network (TSAN) that enables accurate texture sampling from images to preserve detailed facial expressions. To evaluate the proposed method, we conduct extensive experiments on both simulated and real HMC datasets. Experimental results demonstrate that models trained on our synthetic dataset outperform those trained on existing datasets and exhibit better generalization across different camera configurations.

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。

↑