index_path should be "./humanml_index.csv", not "./humanml.csv", at row 181 in file "/mocap-dataset-process/humanml.py"

[Yawningx]

Also at row 30, 31:

smplx_male_model_path = './smplx/SMPLX_MALE.npz'
smplx_female_model_path = './smplx/SMPLX_FEMALE.npz'

Should be:

smplx_male_model_path = './body_models/smplx/SMPLX_MALE.npz'
smplx_female_model_path = './body_models/smplx/SMPLX_FEMALE.npz'

[Yawningx]

While performing face motion augmentation for MOCAP DATA, in:

mv EgoBody_txt ../datasets/texts/semantic_labels/EgoBody
mv humanml_txt ../datasets/texts/semantic_labels/humanml
mv GRAB_txt ../datasets/texts/semantic_labels/GRAB

mv GRAB_txt ../datasets/texts/semantic_labels/GRAB

should be:

mv GRAB_text ../datasets/texts/semantic_labels/GRAB

[whwjdqls]

Hi, thanks for the heads up.
I am wondering how you got the AIST++ dataset.