A Simple Method to Boost Human Pose Estimation Accuracy by Correcting the Joint Regressor for the Human3.6m Dataset
This addresses a specific bottleneck in human pose estimation for computer vision applications, but it is incremental as it refines an existing component rather than introducing a new paradigm.
The paper tackled inaccuracies in the SMPL-to-joint linear layer used in human pose estimation by proposing a method to create pseudo-ground-truth SMPL poses, which improved joint location accuracy and led to better pose estimation results on the Human3.6m test set without retraining.
Many human pose estimation methods estimate Skinned Multi-Person Linear (SMPL) models and regress the human joints from these SMPL estimates. In this work, we show that the most widely used SMPL-to-joint linear layer (joint regressor) is inaccurate, which may mislead pose evaluation results. To achieve a more accurate joint regressor, we propose a method to create pseudo-ground-truth SMPL poses, which can then be used to train an improved regressor. Specifically, we optimize SMPL estimates coming from a state-of-the-art method so that its projection matches the silhouettes of humans in the scene, as well as the ground-truth 2D joint locations. While the quality of this pseudo-ground-truth is challenging to assess due to the lack of actual ground-truth SMPL, with the Human 3.6m dataset, we qualitatively show that our joint locations are more accurate and that our regressor leads to improved pose estimations results on the test set without any need for retraining. We release our code and joint regressor at https://github.com/ubc-vision/joint-regressor-refinement