第一人称视角可见性感知的人体姿态估计
Egocentric Visibility-Aware Human Pose Estimation
针对第一人称视角下关键点不可见问题，提出显式建模可见性信息的方法，并构建带可见性标注的大规模数据集，为遮挡/不可见场景下的姿态估计提供可复用思路与评测基准。
原文依据 none of the existing egocentric HPE datasets provide keypoint visibility annotations
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论使用头戴式设备的第一人称人体姿态估计（HPE）对于各种虚拟现实和增强现实应用至关重要，但由于关键点不可见问题而面临重大挑战。然而，现有的第一人称HPE数据集均未提供关键点可见性标注，且现有方法往往忽略不可见问题，在估计过程中不加区分地处理可见与不可见关键点。因此，它们准确预测可见关键点的能力受到损害。本文首先提出Eva-3M，一个大规模第一人称可见性感知HPE数据集，包含超过300万帧，其中43.5万帧标注了关键点可见性标签。此外，我们为现有的EMHI数据集补充了关键点可见性标注，以进一步促进该方向的研究。进一步，我们提出EvaPose，一种新颖的第一人称可见性感知HPE方法，显式融入可见性信息以提升姿态估计精度。大量实验验证了真实可见性标签在第一人称HPE设定中的重要价值，并表明我们的EvaPose在Eva-3M和EMHI数据集上均达到了最优性能。
英文原文摘要 1221 字符
Egocentric human pose estimation (HPE) using a head-mounted device is crucial for various VR and AR applications, but it faces significant challenges due to keypoint invisibility. Nevertheless, none of the existing egocentric HPE datasets provide keypoint visibility annotations, and the existing methods often overlook the invisibility problem, treating visible and invisible keypoints indiscriminately during estimation. As a result, their capacity to accurately predict visible keypoints is compromised. In this paper, we first present Eva-3M, a large-scale egocentric visibility-aware HPE dataset comprising over 3.0M frames, with 435K of them annotated with keypoint visibility labels. Additionally, we augment the existing EMHI dataset with keypoint visibility annotations to further facilitate the research in this direction. Furthermore, we propose EvaPose, a novel egocentric visibility-aware HPE method that explicitly incorporates visibility information to enhance pose estimation accuracy. Extensive experiments validate the significant value of ground-truth visibility labels in egocentric HPE settings, and demonstrate that our EvaPose achieves state-of-the-art performance in both Eva-3M and EMHI datasets.