search

mks0601 / I2L-MeshNet_RELEASE

Official PyTorch implementation of "I2L-MeshNet: Image-to-Lixel Prediction Network for Accurate 3D Human Pose and Mesh Estimation from a Single RGB Image", ECCV 2020
MIT License
710 stars 130 forks source link

about meta_info['is_3D'] #47

Closed asw91666 closed 3 years ago

asw91666 commented 3 years ago

Could i know why you use meta_info['is_3D']? In model.py code, the meta_info['is_3D'] is used as input of "self.coord_loss" when GT is targets['orig_joint_img']. But i think the code "class CoordLoss" seem no difference whether 'is_3D' is None or not None. Is there anything i know wrong or don't know?

mks0601 commented 3 years ago

All data from datasets with GT 3D pose (e.g., Human3.6M) have is_3D=True. All data from datasets without GT 3D pose (e.g., MSCOCO) have is_3D=False.

asw91666 commented 3 years ago

thank you !!