Abstract
Manual annotation of anatomical landmarks on 3D facial scans is a time-consuming and expertise-dependent task, yet it remains critical for clinical assessments, morphometric analysis, and craniofacial research. While several deep learning methods have been proposed for facial landmark localization, most focus on pseudo-landmarks or require complex input representations, limiting their clinical applicability. This study presents a fully automated deep learning pipeline (PAL-Net) for localizing 50 anatomical landmarks on stereo-photogrammetry facial models. The method combines coarse alignment, region-of-interest filtering, and an initial approximation of landmarks with a patch-based pointwise CNN enhanced by attention mechanisms. Trained and evaluated on 214 annotated scans from healthy adults, PAL-Net achieved a mean localization error of 3.686 mm and preserves relevant anatomical distances with a 2.822 mm average error, comparable to intra-observer variability. To assess generalization, the model was further evaluated on 700 subjects from the FaceScape dataset, achieving a point-wise error of 0.41\,mm and a distance-wise error of 0.38\,mm. Compared to existing methods, PAL-Net offers a favorable trade-off between accuracy and computational cost. While performance degrades in regions with poor mesh quality (e.g., ears, hairline), the method demonstrates consistent accuracy across most anatomical regions. PAL-Net generalizes effectively across datasets and facial regions, outperforming existing methods in both point-wise and structural evaluations. It provides a lightweight, scalable solution for high-throughput 3D anthropometric analysis, with potential to support clinical workflows and reduce reliance on manual annotation. Source code can be found at this https URL
Abstract (translated)
三维面部扫描的解剖标志的手动注释是一项耗时且依赖专业知识的任务，但对于临床评估、形态学分析和颅面研究来说至关重要。尽管已经提出了一些深度学习方法用于面部特征定位，但大多数这些方法要么关注伪标志，要么需要复杂的输入表示，这限制了它们在临床上的应用。这项研究提出了一种全自动的深度学习流水线（PAL-Net），用于定位立体摄影测量面部模型上的50个解剖标志。该方法结合了粗略对齐、感兴趣区域过滤以及基于注意力机制增强的补丁级点卷积神经网络，以实现初始特征位置估计。 在214张来自健康成年人的注释扫描上训练和评估后，PAL-Net实现了3.686毫米的平均定位误差，并且在保留相关解剖距离方面达到了2.822毫米的平均误差，这一成绩与观察者内部变异性的水平相当。为了评估模型的一般化能力，研究团队进一步使用FaceScape数据集中700个受试者的面部扫描对PAL-Net进行了测试，结果显示点间误差为0.41毫米，结构距离误差为0.38毫米。 相比于现有的方法，PAL-Net在精度和计算成本之间提供了更优的权衡。尽管在网格质量较差的区域（如耳朵、发际线）表现有所下降，但该方法在大多数解剖区域内均表现出稳定的准确性。此外，PAL-Net能够有效地跨数据集和面部区域进行泛化，并且在点间和结构评估方面优于现有方法。 这项研究为大规模高通量3D人体测量分析提供了一个轻量化、可扩展的解决方案，有可能支持临床工作流程并减少对手动注释的依赖。该研究的相关源代码可以在提供的链接中找到（此处未给出具体网址）。
URL
https://arxiv.org/abs/2510.00910
