论文标题
MEEV：以自我为中心视频的身体网格估计
MEEV: Body Mesh Estimation On Egocentric Video
论文作者
论文摘要
该技术报告介绍了我们的解决方案MEEV，向ECCV 2022的Egobody挑战提出了建议。该数据集由头部安装的设备捕获，数据集由人体形状和相互作用的人的运动组成。 Egobody数据集具有遮挡的身体或模糊图像等挑战。为了克服挑战，MEEV旨在利用多尺度功能来获得丰富的空间信息。此外，为了克服有限的数据集大小，该模型已通过数据集汇总的2D和3D姿势估计数据集进行了预训练。 MEEV在MPJPE中获得82.30，MPVPE达到92.93，在ECCV 2022上赢得了Egobody挑战，该挑战显示了所提出的方法的有效性。该代码可从https://github.com/clovaai/meev获得
This technical report introduces our solution, MEEV, proposed to the EgoBody Challenge at ECCV 2022. Captured from head-mounted devices, the dataset consists of human body shape and motion of interacting people. The EgoBody dataset has challenges such as occluded body or blurry image. In order to overcome the challenges, MEEV is designed to exploit multiscale features for rich spatial information. Besides, to overcome the limited size of dataset, the model is pre-trained with the dataset aggregated 2D and 3D pose estimation datasets. Achieving 82.30 for MPJPE and 92.93 for MPVPE, MEEV has won the EgoBody Challenge at ECCV 2022, which shows the effectiveness of the proposed method. The code is available at https://github.com/clovaai/meev
