VideoRecognition/helperFunctions.py at master · udianand/VideoRecognition · 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
import os
import numpy as np
import cv2
import time
import h5py

mean_subtract = np.asarray([103.062623801, 115.902882574, 123.151630838],np.float32)

curr_w = 320
curr_h = 240
height = width = 224

def loadFrame(args):
    (filename,augment) = args

    data = np.zeros((height,width,3),dtype=np.float32)

    try:
        ### load file from HDF5
        filename = filename.replace('.avi','.hdf5')
        filename = filename.replace('UCF-101','UCF-101-hdf5')
        h = h5py.File(filename,'r')
        nFrames = len(h['video'])
        frame_index = np.random.randint(nFrames)
        frame = h['video'][frame_index]

        if(augment==True):
            ## RANDOM CROP - crop 70-100% of original size
            ## don't maintain aspect ratio
            if(np.random.randint(2)==0):
                resize_factor_w = 0.3*np.random.rand()+0.7
                resize_factor_h = 0.3*np.random.rand()+0.7
                w1 = int(curr_w*resize_factor_w)
                h1 = int(curr_h*resize_factor_h)
                w = np.random.randint(curr_w-w1)
                h = np.random.randint(curr_h-h1)
                frame = frame[h:(h+h1),w:(w+w1)]

            ## FLIP
            if(np.random.randint(2)==0):
                frame = cv2.flip(frame,1)

            frame = cv2.resize(frame,(width,height))
            frame = frame.astype(np.float32)

            ## Brightness +/- 15
            brightness = 30
            random_add = np.random.randint(brightness+1) - brightness/2.0
            frame += random_add
            frame[frame>255] = 255.0
            frame[frame<0] = 0.0

            ## shrink image by up to 10% within frame
            shrink_size =  int(0.10*np.random.rand()*width)
            frame = cv2.copyMakeBorder(frame,top=shrink_size,bottom=shrink_size,left=shrink_size,right=shrink_size,borderType=cv2.BORDER_CONSTANT,value=(0.0, 0.0, 0.0))
            frame = cv2.resize(frame, (width,height))

            ## random rotate  +/- 12.5 degrees
            angle = 25
            random_rotate = np.random.randint(angle+1) - angle/2
            M = cv2.getRotationMatrix2D((width/2,height/2),random_rotate,1.0)
            frame = cv2.warpAffine(frame,M,(width,height),borderMode=cv2.BORDER_CONSTANT,borderValue=(0.0, 0.0, 0.0))
        else:
            # don't augment
            frame = cv2.resize(frame,(width,height))
            frame = frame.astype(np.float32)

        ## resnet model was trained on images with mean subtracted
        data[:,:,:] = frame - mean_subtract
    except:
        print("Exception: " + filename)
        data = np.array([])
    return data

def getUCF101(base_directory = ''):

    # action class labels
    class_file = open('ucfTrainTestlist/classInd.txt','r')
    lines = class_file.readlines()
    lines = [line.split(' ')[1].strip() for line in lines]
    class_file.close()
    class_list = np.asarray(lines)

    # training data
    train_file = open('ucfTrainTestlist/trainlist01.txt','r')
    lines = train_file.readlines()
    filenames = ['UCF-101/' + line.split(' ')[0] for line in lines]
    y_train = [int(line.split(' ')[1].strip())-1 for line in lines]
    y_train = np.asarray(y_train)
    filenames = [base_directory + filename for filename in filenames]
    train_file.close()

    train = (np.asarray(filenames),y_train)

    # testing data
    test_file = open('ucfTrainTestlist/testlist01.txt','r')
    lines = test_file.readlines()
    filenames = ['UCF-101/' + line.split(' ')[0].strip() for line in lines]
    classnames = [filename.split('/')[1] for filename in filenames]
    y_test = [np.where(classname == class_list)[0][0] for classname in classnames]
    y_test = np.asarray(y_test)
    filenames = [base_directory + filename for filename in filenames]
    test_file.close()

    test = (np.asarray(filenames),y_test)

    return class_list, train, test

def loadSequence(args):

    (filename,sequence_length,is_train) = args

    try:
        filename = filename.replace('.avi','.hdf5')
        filename = filename.replace('UCF-101','UCF-101-features')

        h = h5py.File(filename,'r')
        nFrames = len(h['features'])

        if(nFrames<sequence_length):
            start_index = 0
            new_sequence_length = int(nFrames)
        else:
            start_index = np.random.randint(nFrames - (sequence_length-1))
            new_sequence_length = sequence_length

        indices = range(start_index, start_index+new_sequence_length)
        features = h['features'][indices]
        h.close()

        if(features.shape[0]<sequence_length): # didn't have enough frames in this video to fully load it
            features = repeatSequence(features,sequence_length)
            features = features[0:sequence_length]
        if(is_train):
            if(np.random.randint(2)==0):
                features = np.flipud(features)

    except:
        print("Exception: " + filename)
        features = np.array([])

    return features

def repeatSequence(rnn_input,length_of_sequence):

    while(rnn_input.shape[0]<length_of_sequence):
        rnn_input_flipped = np.flipud(rnn_input)
        rnn_input=np.vstack((rnn_input,rnn_input_flipped))
    return rnn_input