Human Pose Estimation from Sparse Inertial Measurements through Recurrent Graph Convolution

Conventional methods for human pose estimation either require a high degree of instrumentation, by relying on many inertial measurement units (IMUs), or constraint the recording space, by relying on extrinsic cameras. These deficits are tackled through the approach of human pose estimation from sparse IMU data. We define adjacency adaptive graph convolutional long-short term memory networks (AAGC-LSTM), to tackle human pose estimation based on six IMUs, while incorporating the human body graph structure directly into the network. The AAGC-LSTM combines both spatial and temporal dependency in a single network operation, more memory efficiently than previous approaches. This is made possible by equipping graph convolutions with adjacency adaptivity, which eliminates the problem of information loss in deep or recurrent graph networks, while it also allows for learning unknown dependencies between the human body joints. To further boost accuracy, we propose longitudinal loss weighting to consider natural movement patterns. With our presented approach, we are able to utilize the inherent graph nature of the human body, and thus can outperform the state of the art (SOTA) for human pose estimation from sparse IMU data.