catalina17 / xflow Goto Github PK

Hi Catalina, it is glad to see your codes. Wonderful. But I have a few questions.
1 digits_data.py What is Y nad P? and why you set Y = np.empty(shape=(total,), dtype=np.int) and P = np.empty(shape=(total,), dtype=np.dtype('a16')) as an array.
2 Currently, I want to use your code for audio and image classification. How could I update code in digits_data.py? Because you use sliding window for short video? many thanks.

Hi,

I would like to know how were min_frames and max-frames for determined as if I count the number of frames for the Digits dataset you supply in your GitHub, I found a min of 7 and a max of 63 when in the code you present in the git, you have them respectively at 6 and 58.
Thanks in advance for your answer

Hi,

May I know how you get the MFCC features? Like the parameters you used. Thank you very much!