Closed zhengthomastang closed 3 years ago
Hi! The preprocess_XXX.py already includes the preprocessing step (l. 143, fp = pre_normalization(fp)). If you instead use XXX_gendata.py, you should run the pre-normalization step on the generated data.
Chiara
For the NTU RGB+D datasets, there are
XXX_gendata.py
andpreprocess_XXX.py
which read the same inputs and produce the same outputs. What are their differences and how the results differ if training on the corresponding.npy
and.label
files generated?