KeyIdentification/main.py at master · RajSSS5/KeyIdentification · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
# A Wavenet For Source Separation - Francesc Lluis - 25.10.2018
# Extended in 2019 - Sukhraj Sekhon
# Main.py

import sys
import logging
import optparse
import json
import os
import models
import jams
import scipy
import numpy as np

from datasets import GiantStepsDataset
import util


encoded_key_vals = np.eye(24)
encoded_key_vals = [tuple(arr) for arr in encoded_key_vals]

key_vals = {
    encoded_key_vals[0]:'A:minor',
    encoded_key_vals[1]:'A:major',
    encoded_key_vals[2]:'Ab:minor',
    encoded_key_vals[3]:'Ab:major',
    encoded_key_vals[4]:'B:minor',
    encoded_key_vals[5]:'B:major',
    encoded_key_vals[6]:'Bb:minor',
    encoded_key_vals[7]:'Bb:major',
    encoded_key_vals[8]:'C:minor',
    encoded_key_vals[9]:'C:major',
    encoded_key_vals[10]:'D:minor',
    encoded_key_vals[11]:'D:major',
    encoded_key_vals[12]:'Db:minor',
    encoded_key_vals[13]:'Db:major',
    encoded_key_vals[14]:'E:minor',
    encoded_key_vals[15]:'E:major',
    encoded_key_vals[16]:'Eb:minor',
    encoded_key_vals[17]:'Eb:major',
    encoded_key_vals[18]:'F:major',
    encoded_key_vals[19]:'F:minor',
    encoded_key_vals[20]:'G:minor',
    encoded_key_vals[21]:'G:major',
    encoded_key_vals[22]:'Gb:minor',
    encoded_key_vals[23]:'Gb:major'

}

trainer = GiantStepsDataset("train")
tester = GiantStepsDataset("test")


def load_config(config_filepath):
    try:
        config_file = open(config_filepath, 'r')
    except IOError:
        logging.error('No readable config file at path: ' + config_filepath)
        exit()
    else:
        with config_file:
            return json.load(config_file)
input_len = 500000

def training(config_letter, config_key_type, cla):

    #Instantiate model

    # model_letter = models.WavenetClassifier(config_letter, load_checkpoint=cla.load_checkpoint, input_length=input_length)
    model_key_type = models.WavenetClassifier(config_key_type, load_checkpoint=cla.load_checkpoint, input_length=input_len)

    num_steps_train_key_type = config_key_type['training']['num_steps_train']
    num_steps_test_key_type = config_key_type['training']['num_steps_test']

    batch_train_key_type = trainer.load_dataset_generator()
    batch_test_key_type = tester.load_dataset_generator()

    # batch = {'data_input': sequences}, {'letters': letters, 'key_types': batch_outputs_2}

    model_key_type.fit_model(batch_train_key_type, num_steps_train_key_type, batch_test_key_type, num_steps_test_key_type, config_key_type['training']['num_epochs'])
    # model_letter.fit_model(batch_train_letter, num_steps_train_letter, batch_test_letter, num_steps_test_letter, config_letter['training']['num_epochs'])

def get_command_line_arguments():
    parser = optparse.OptionParser()
    parser.set_defaults(mode='training')
    parser.set_defaults(load_checkpoint=None)
    parser.set_defaults(source=None)

    parser.add_option('--mode', dest='mode')
    parser.add_option('--load_checkpoint', dest='load_checkpoint')

    parser.add_option('--source', dest='source')

    (options, args) = parser.parse_args()

    return options


def inference(config_letter, config_key_type, cla):
    model_key_type = models.WavenetClassifier(config_key_type, load_checkpoint=cla.load_checkpoint, input_length=input_len)

    if cla.source.endswith('.wav'):
        filenames = [cla.source.rsplit('/', 1)[-1]]
        cla.source = cla.source.rsplit('/', 1)[0] + '/'

    else:
        if not cla.source.endswith('/'):
            cla.source += '/'
        filenames = [filename for filename in os.listdir(cla.source) if filename.endswith('.wav')]


    jams_directory = r"dataset\annotations\jams"

    for filename in filenames:
        inp = util.load_wav(cla.source + filename, 16000)
        inp = inp[:input_len]
        f, t, Sxx = scipy.signal.spectrogram(inp, 16000)
        Sxx = np.expand_dims(Sxx, axis=-1)
        Sxx = np.expand_dims(Sxx, axis=0)


        jams_name = filename[:-4] + ".jams"
        jams_path = os.path.join(jams_directory, jams_name)
        jamsObj = jams.load(jams_path)
        data_val = jamsObj.annotations[0].data[0].value

        result = key_vals[tuple(model_key_type.predict(Sxx)[0].astype(int))]
        print('File: ', filename, 'Prediction: ', result, 'Actual: ', data_val)


def set_system_settings():
    sys.setrecursionlimit(50000)
    logging.getLogger().setLevel(logging.INFO)

def get_valid_output_folder_path(outputs_folder_path):
    j = 1
    while True:
        output_folder_name = 'samples_%d' % j
        output_folder_path = os.path.join(outputs_folder_path, output_folder_name)
        if not os.path.isdir(output_folder_path):
            os.mkdir(output_folder_path)
            break
        j += 1
    return output_folder_path

def main():
    set_system_settings()
    cla = get_command_line_arguments()

    config_letter = load_config('letter_config.json')
    config_key_type = load_config('key_type_config.json')

    if cla.mode == 'training':
        training(config_letter, config_key_type, cla)
    elif cla.mode == 'inference':
        inference(config_letter, config_key_type, cla)


if __name__ == "__main__":
    main()