View Invariant Human Body Detection and Pose Estimation from Multiple Depth Sensors
This addresses occlusion problems in indoor monitoring applications such as hospital operation rooms, offering a more efficient fusion method compared to complex models.
The paper tackles multi-person 3D pose estimation in indoor monitoring scenarios using multiple depth sensors, proposing Point R-CNN which concatenates point clouds at the input level and outperforms cascaded state-of-the-art models in experiments on datasets like CMU panoptic and MVOR.
Point cloud based methods have produced promising results in areas such as 3D object detection in autonomous driving. However, most of the recent point cloud work focuses on single depth sensor data, whereas less work has been done on indoor monitoring applications, such as operation room monitoring in hospitals or indoor surveillance. In these scenarios multiple cameras are often used to tackle occlusion problems. We propose an end-to-end multi-person 3D pose estimation network, Point R-CNN, using multiple point cloud sources. We conduct extensive experiments to simulate challenging real world cases, such as individual camera failures, various target appearances, and complex cluttered scenes with the CMU panoptic dataset and the MVOR operation room dataset. Unlike most of the previous methods that attempt to use multiple sensor information by building complex fusion models, which often lead to poor generalization, we take advantage of the efficiency of concatenating point clouds to fuse the information at the input level. In the meantime, we show our end-to-end network greatly outperforms cascaded state-of-the-art models.