This pull request implements the per_keypoint_evaluation and comparisonbodyparts parameters for PyTorch pose estimation models.

Updates

• New per_keypoint_evaluation argument for deeplabcut.pose_estimation_pytorch.apis.evaluation.evaluate_network to compute the errors (RMSE and mAP) based on a subset of bodyparts, as done for tensorflow models
• New comparisonbodyparts argument for deeplabcut.pose_estimation_pytorch.apis.evaluation.evaluate_network to compute RMSE per-bodypart and store the results in a CSV file, as done for tensorflow models
• Inclusion of unique bodypart RMSE in the "main" RMSE computation to match the tensorflow RMSE computation
• Additional tests to verify that RMSE computation is correct
• Updates the functional PyTorch tests to compute per-bodypart RMSE
• Update docs/pytorch/user_guide.md so that methods where parameters still need to be implemented are labeled orange and not green
• Update docs/pytorch/user_guide.md to reflect that comparisonbodyparts and per_keypoint_evaluation can now be used with PyTorch models

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.