A speech emotion recognition (SER) system that won 1st place in a competition for a deep learning master-course. It uses a parallel CNN built using Keras and Tensorflow.
The Google collab notebooks parallel-CNN.ipynb and Parallel-CNN-self-validate.ipynb provided in the source are corrupt.
Not able to open the files in Google collab.
Please provide valid notebooks.
I ran parallel-CNN.py on Windows10
but I had a problem.....
(tfpose) D:\cap\SER-Parallel-CNN>python parallel-CNN.py
Traceback (most recent call last):
File "parallel-CNN.py", line 79, in
logMel_train_ids,logMel_train_features,logMel_train_labels = unpickle_data(logMel_file_paths[0])
File "parallel-CNN.py", line 70, in unpickle_data
features = pickle.load(f_in)
UnicodeDecodeError: 'ascii' codec can't decode byte 0xc6 in position 2: ordinal not in range(128)