Abstract: We propose a multi-view weakly-supervised 3D human pose estimation network for depth maps. In this network, we introduce the Sum of Gaussians (SoG) model to represent the human body and ...

The figure below shows an overview of the different MiDaS models; the bubble size scales with number of parameters. The argument --side is optional and causes both the input RGB image and the output ...

This study addresses this challenge by proposing a novel, automated pipeline for generating labelled synthetic multiclass pothole depth-map datasets, simulating data captured by a depth camera. This ...