MultipleReservoirComputing-MRC/LoadExtractedKeypoint.py at main · tamukohlaboratory/MultipleReservoirComputing-MRC · 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
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
import numpy as np
import os
import os.path
# from tslearn.preprocessing import TimeSeriesResampler
# import tensorflow as tf


class LoadExtractedKeypoint:
    maxFrame = 0

    def setMaxFrame(self, lenghtFrame=0):
        if (lenghtFrame > self.maxFrame):
            self.maxFrame = lenghtFrame

    def getMaxFrame(self):
        return self.maxFrame

    def LoadKeypoint(self, pathWorkspace, folder_path_keypoint,
                     file_name_data_numpy, file_name_label_numpy):
        xTemp = []
        newXTemp = []
        yTemp = []
        DATA_PATH_KEYPOINT = os.path.join(pathWorkspace, folder_path_keypoint)
        actionsVideoInput = np.array(os.listdir(DATA_PATH_KEYPOINT))

        for action in actionsVideoInput:
            print("action = {}".format(action))
            for actionSubFolder in np.array(os.listdir(os.path.join(DATA_PATH_KEYPOINT, action))):
                sequences = 1
                window = []
                tempSequencesWhichExist = 0
                # list all npy file for each frame in the video
                allKeyPointData = sorted(os.listdir(os.path.join(DATA_PATH_KEYPOINT,
                                                          action,
                                                          actionSubFolder)))
                print(f'{allKeyPointData =}')
                for _ in range(len(allKeyPointData)):
                    print(f'{_}')
                    pathFile = os.path.join(DATA_PATH_KEYPOINT, action,
                                            actionSubFolder,
                                            "{}.npy".format(sequences))

                    print(f'{pathFile=}')

                    # if file not exist take the previous keypoint
                    if (os.path.isfile(pathFile) is False):
                        pathFile = os.path.join(DATA_PATH_KEYPOINT, action,
                                                actionSubFolder,
                                                "{}.npy".format(tempSequencesWhichExist))
                    else:
                        tempSequencesWhichExist = sequences

                    res = np.load(pathFile)
                    window.append(res)
                    sequences += 1

                # ### interpolation in here
                # window        = np.asarray(window, dtype=object)
                # array_reshape = np.reshape(window, window.shape[0]* window.shape[1])
                # size          = 78 * window.shape[1]
                # result_interpolation  = TimeSeriesResampler(sz = size).fit_transform(array_reshape)
                # final_window  = np.reshape(result_interpolation, (78, window.shape[1]))
                # window = np.array(window,dtype=object)
                # final_window  = np.transpose(window,[1,0])
                final_window = np.asarray(window)
                # print (f'{np.asarray(final_window).shape=}')

                self.setMaxFrame(lenghtFrame=final_window.shape[0])
                xTemp.append(final_window)
                yTemp.append(action)

        # process padding
        print(f'{self.getMaxFrame()=}')
        print(len(xTemp))
        # xTemp = np.asarray(xTemp)
        for jj in range(len(xTemp)):
            # to make sure if shape xTemp 2D
            if len(xTemp[jj].shape) == 3:
                xTemp[jj] = xTemp[jj].reshape((xTemp[jj].shape[0], -1))
            tempTranspose = np.transpose(xTemp[jj], [1, 0])
            # print(f'Shape after transpose: {tempTranspose.shape}')
            lenghtPadding = self.getMaxFrame() - tempTranspose.shape[1]
            if lenghtPadding < 0:
                # print(f"Warning: Length padding is negative for index {jj}, array shape {tempTranspose.shape}")
                continue
            paddResult = np.pad(tempTranspose, ((0, 0), (0, lenghtPadding)), constant_values=0)
            paddResult = np.transpose(paddResult, [1, 0])
            # print(f'Shape after padding: {paddResult.shape}')
            newXTemp.append(paddResult)

        print(f'{self.getMaxFrame()=}')
        # save keypoint
        pathKeypoint = os.path.join(pathWorkspace, 'DataSaveOnNumpy',
                                    file_name_data_numpy)
        # np.save (pathKeypoint, xTemp)
        np.save(pathKeypoint, newXTemp)
        print(np.asarray(newXTemp).shape)
        pathLabel = os.path.join(pathWorkspace, 'DataSaveOnNumpy',
                                 file_name_label_numpy)
        np.save(pathLabel, yTemp)


if __name__ == "__main__":
    PATH_WORKSPACE = '/home/bra1n/Documents/signLanguage/MRC'

    # 100 normalize
    detail_filename = 'normalization_option2' # 'without_normalization_option1'   # 'normalization_option2'  # 'without_normalization_option1'
    FOLDER_PATH_TRAINING = "Keypoint{}WLASL100_{}".format("Training", detail_filename)
    FOLDER_PATH_VALIDATION = "Keypoint{}WLASL100_{}".format("Validation", detail_filename)
    FOLDER_PATH_TESTING = "Keypoint{}WLASL100_{}".format("Testing", detail_filename)

    FILE_NAME_TRAINING_NUMPY = "{}AllFrame_WLASL_100Class_{}".format("Training", detail_filename)
    FILE_NAME_LABEL_TRAINING_NUMPY = "{}AllFrame_WLASL_100Class_{}".format("TrainingLabel", detail_filename)

    FILE_NAME_VALIDATION_NUMPY = "{}AllFrame_WLASL_100Class_{}".format("Validation", detail_filename)
    FILE_NAME_LABEL_VALIDATION_NUMPY = "{}AllFrame_WLASL_100Class_{}".format("ValidationLabel", detail_filename)

    FILE_NAME_TEST_NUMPY = "{}AllFrame_WLASL_100Class_{}".format("Testing", detail_filename)
    FILE_NAME_LABEL_TEST_NUMPY = "{}AllFrame_WLASL_100Class_{}".format("TestingLabel", detail_filename)

    # 100 without
    # FOLDER_PATH_TRAINING = "Keypoint{}WLASL100_without_normalization_option1".format("Training")
    # FOLDER_PATH_VALIDATION = "Keypoint{}WLASL100_without_normalization_option1".format("Validation")
    # FOLDER_PATH_TESTING = "Keypoint{}WLASL100_without_normalization_option1".format("Testing")

    # FILE_NAME_TRAINING_NUMPY = "{}AllFrame_WLASL_100Class_without_option1".format("Training")
    # FILE_NAME_LABEL_TRAINING_NUMPY = "{}AllFrame_WLASL_100Class_without_option1".format("TrainingLabel")

    # FILE_NAME_VALIDATION_NUMPY = "{}AllFrame_WLASL_100Class_without_option1".format("Validation")
    # FILE_NAME_LABEL_VALIDATION_NUMPY = "{}AllFrame_WLASL_100Class_without_option1".format("ValidationLabel")

    # FILE_NAME_TEST_NUMPY = "{}AllFrame_WLASL_100Class_without_option1".format("Testing")
    # FILE_NAME_LABEL_TEST_NUMPY = "{}AllFrame_WLASL_100Class_without_option1".format("TestingLabel")

    # # 300 without
    # FOLDER_PATH_TRAINING = "Keypoint{}WLASL300_without_normalization_option1".format("Training")
    # FOLDER_PATH_VALIDATION = "Keypoint{}WLASL300_without_normalization_option1".format("Validation")
    # FOLDER_PATH_TESTING = "Keypoint{}WLASL300_without_normalization_option1".format("Testing")

    # FILE_NAME_TRAINING_NUMPY = "{}AllFrame_WLASL_300Class_without_option1".format("Training")
    # FILE_NAME_LABEL_TRAINING_NUMPY = "{}AllFrame_WLASL_300Class_without_option1".format("TrainingLabel")

    # FILE_NAME_VALIDATION_NUMPY = "{}AllFrame_WLASL_300Class_without_option1".format("Validation")
    # FILE_NAME_LABEL_VALIDATION_NUMPY = "{}AllFrame_WLASL_300Class_without_option1".format("ValidationLabel")

    # FILE_NAME_TEST_NUMPY = "{}AllFrame_WLASL_300Class_without_option1".format("Testing")
    # FILE_NAME_LABEL_TEST_NUMPY = "{}AllFrame_WLASL_300Class_without_option1".format("TestingLabel")

    FOLDER_SAVE_NUMPY = os.path.join(PATH_WORKSPACE, 'DataSaveOnNumpy')
    if not os.path.exists(FOLDER_SAVE_NUMPY):
        os.makedirs(FOLDER_SAVE_NUMPY)

    sequencesTraining = []
    labelsTraining = []
    sequencesTesting = []
    labelsTesting = []

    tempLoadExtractedKeypoint = LoadExtractedKeypoint()

    tempLoadExtractedKeypoint.LoadKeypoint(
                                PATH_WORKSPACE,
                                FOLDER_PATH_TRAINING,
                                FILE_NAME_TRAINING_NUMPY,
                                FILE_NAME_LABEL_TRAINING_NUMPY)

    tempLoadExtractedKeypoint.LoadKeypoint(
                                PATH_WORKSPACE,
                                FOLDER_PATH_VALIDATION,
                                FILE_NAME_VALIDATION_NUMPY,
                                FILE_NAME_LABEL_VALIDATION_NUMPY)

    tempLoadExtractedKeypoint.LoadKeypoint(
                                PATH_WORKSPACE,
                                FOLDER_PATH_TESTING,
                                FILE_NAME_TEST_NUMPY,
                                FILE_NAME_LABEL_TEST_NUMPY)