kcct-fujimotolab / 3DCNN

3D convolutional neural network for video classification
270 stars 105 forks source link

data format to network #21

Open lmhlmh1282 opened 4 years ago

lmhlmh1282 commented 4 years ago

why is the format of the data like (n,img_rows, img_cols, frames, channel) rather than (n,f,rows,cols,channel) ?