The repository that this is forked from 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! This repo aims to extend the capabilites to the CMU dataset
hi, I want to know what is the coco 19 format. Could you give an example？ https://github.com/Samleo8/learnable-triangulation-pytorch/blob/master/SETUP_GENERAL_DATASET.md#pose-data-needed-for-training-and-volumetric-triangulation-testing