Traditional 2D pose estimation models are limited by their category-specific
design, making them suitable only for predefined object categories. This
restriction becomes particularly challenging when dealing with novel objects
due to the lack of relevant training data.
To address this limitation, category-agnostic pose estimation (CAPE) was
introduced. CAPE aims to enable keypoint localization for arbitrary object
categories using a single model, requiring minimal support images with
annotated keypoints. This approach not only enables object pose generation
based on arbitrary keypoint definitions but also significantly reduces the
associated costs, paving the way for versatile and adaptable pose estimation
applications.
We present a novel approach to CAPE that leverages the inherent geometrical
relations between keypoints through a newly designed Graph Transformer Decoder.
By capturing and incorporating this crucial structural information, our method
enhances the accuracy of keypoint localization, marking a significant departure
from conventional CAPE techniques that treat keypoints as isolated entities.
We validate our approach on the MP-100 benchmark, a comprehensive dataset
comprising over 20,000 images spanning more than 100 categories. Our method
outperforms the prior state-of-the-art by substantial margins, achieving
remarkable improvements of 2.16% and 1.82% under 1-shot and 5-shot settings,
respectively. Furthermore, our method's end-to-end training demonstrates both
scalability and efficiency compared to previous CAPE approaches.

传统的 2D 姿态估计模型局限于其特定类别的设计，限定了其适用范围，对于新颖对象缺乏相关训练数据的情况尤其具有挑战性。为了应对这一局限性，引入了无类别限制的姿态估计（CAPE）方法，在仅需标注关键点的最小支持图像的情况下，实现了任意对象类别的关键点定位。我们提出了一种利用新设计的图转换解码器的 CAPE 新方法，通过捕捉和整合关键点之间固有的几何关系信息，进一步提高了关键点定位的准确性，与传统 CAPE 技术将关键点视为孤立实体的方式存在显著区别。我们在包含超过 100 个类别的 20,000 多张图像的 MP-100 基准数据集上验证了我们的方法，与之前最先进的方法相比，本方法在 1-shot 和 5-shot 设置下均取得了显著的改进，分别达到了 2.16％和 1.82％的提升。此外，与之前的 CAPE 方法相比，我们的方法的端对端训练表现出良好的可扩展性和效率。