Closed snorlaxse closed 2 years ago
Hi SnorlaxSE, sorry for the confusion. We used "27" and "27_2" as two groups of keypoints (w/ or w/o test-time augmentation) when extracting the whole-body poses. But now they are the same thing. I have updated the code in Commit de0c9ac.
Thanks for your reply. And I have another question where ensemble/gcn/*.pkls comes from.
Thanks for your reply. And I have another question where ensemble/gcn/*.pkls comes from.
The ensemble/gcn/*.pkls are testing results of SL-GCN (on AUTSL dataset). You can use our provided pretrained models and preprocessed skeleton data to obtain those results.
https://github.com/jackyjsy/CVPR21Chal-SLR/blob/892de189f4f8afc69031780a44f308044eff2429/SL-GCN/data_gen/sign_gendata.py#L15
The '27' and '27_2' info of selected_joints is the same.
Could you please let me know what the idea is?