摘要：尽管人脸关键点检测（FLD）方法的进展持续推动性能边界，但它们忽视了两项主要的功能局限：（1）针对每个“N 点”基准数据集，都需要独立训练不同的网络参数；（2）在“N 点”数据集上训练的模型，通常只能可靠地输出这 $N$ 个关键点。在本文中，我们首先提出“人脸部位锚定关键点位置”（Face Part-Anchored Landmark Positions，FPALPs）的概念：每个关键点均被视为沿某个人脸部位轮廓、从零（起点）到一（终点）的进度值。无论关键点来自哪个源数据集，每个关键点都可以用 FPALP 格式表示，因此能够将所有“N 点”数据集统一为单一数据集。其次，我们使用基于 FPALP 的查询表示每个关键点，通过跨模态解码器对其逐步细化，并依据最终表示预测其坐标。我们提出的方法名为 Unified Dynamic FLD，体现了这两项设计选择，并通过以下能力简化了关键点检测流程：（1）单一模型可在任意数量的“N 点”数据集上学习；（2）在运行时加载指定的关键点查询，即可生成任意数量的特定关键点预测。在多个基准数据集上的广泛实验表明，我们的方法在具备这些优势的同时，仍能与现有最先进方法保持竞争力，并在若干情况下超越它们。

Abstract:Although advancements in face landmark detection (FLD) methods continue to push performance boundaries, they overlook two major functional limitations: (1) different network parameters need to be trained independently for each $N$-point'' benchmark dataset, and (2) a model trained on an $N$-point'' dataset reliably outputs only the $N$ landmarks. In our work, we first conceptualize Face Part-Anchored Landmark Positions (FPALPs), wherein each landmark is treated as a progression value between zero (start) and one (end) along a face part's contour. Every landmark can be expressed in the FPALP format, irrespective of its source dataset, hence unlocking the ability to unify all $N$-point'' datasets into a single dataset. Secondly, we represent each landmark with an FPALP-based query, refine it progressively with a cross-modality decoder, and predict its coordinates based on the final representation. Our approach, called Unified Dynamic FLD, embodies these two design choices and streamlines the landmark detection pipeline by enabling (1) a single model to learn on any number of $N$-point'' datasets, and (2) yield any number of specific landmark predictions by loading the designated landmark queries at runtime. Extensive experiments on multiple benchmark datasets show that our method delivers these benefits while remaining competitive with, and in several cases outperforming existing state-of-the-art methods.