This repository is an official PyTorch implementation of the paper "Learnable Triangulation of Human Pose" (ICCV 2019, oral). Proposed method archives state-of-the-art results in multi-view 3D human pose estimation!
Hello! Thanks for your work!
I am writing to request access to the model trained on the CMU Panoptic dataset.I encountered a problem that a model trained on Human36 performs poorly on CMU with a large number of occlusions
Hello! Thanks for your work! I am writing to request access to the model trained on the CMU Panoptic dataset.I encountered a problem that a model trained on Human36 performs poorly on CMU with a large number of occlusions