CNN-handwritten-digits-recognition/test_cnn.py at main · Regen2001/CNN-handwritten-digits-recognition · 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
from __future__ import print_function
import argparse
import torch
import torch.nn as nn
import torch.nn.functional as F
from torchvision import datasets, transforms
import matplotlib.pyplot as plt
import os
import numpy as np
import cv2
import matplotlib.pyplot as plt
import shutil
import random


class Net(nn.Module):
    def __init__(self):
        super(Net, self).__init__()
        self.conv1 = nn.Conv2d(1, 8, 3, 1)
        self.conv2 = nn.Conv2d(8, 16, 3, 1)
        self.dropout1 = nn.Dropout(0.25)
        self.dropout2 = nn.Dropout(0.5)
        self.fc1 = nn.Linear(2304, 64)
        self.fc2 = nn.Linear(64, 10)

    def forward(self, x):
        x = self.conv1(x)
        x = F.relu(x)
        x = self.conv2(x)
        x = F.relu(x)
        x = F.max_pool2d(x, 2)
        x = self.dropout1(x)
        x = torch.flatten(x, 1)
        x = self.fc1(x)
        x = F.relu(x)
        x = self.dropout2(x)
        x = self.fc2(x)
        return x

def main(args):

    device = torch.device("cuda" if (torch.cuda.is_available() and args.use_cuda) else "cpu")

    model = Net().to(device)
    model.load_state_dict(torch.load("best_model/mnist_cnn.pt"))

    fileCreate('pred_res')
    fileCreate('pred_res/correct')
    fileCreate('pred_res/false')

    test_loader = dataset(args)

    test(model, device, test_loader)

    return 0

def parse_args():
    parser = argparse.ArgumentParser(description='PyTorch MNIST Example')

    parser.add_argument('--test-batch-size', type=int, default=1000, metavar='N', help='input batch size for testing (default: 1000)')

    parser.add_argument('--num-workers', type=int, default=4, metavar='N', help='number of worker of torch to train (default: 4)')

    parser.add_argument('--use-cuda', action='store_true', default=True, help='disables CUDA training')

    return parser.parse_args()

def dataset(args):

    test_kwargs = {'batch_size': args.test_batch_size}
    if torch.cuda.is_available() and args.use_cuda:
        cuda_kwargs = {'num_workers': args.num_workers,
                        'pin_memory': True,
                        'shuffle': True}
    test_kwargs.update(cuda_kwargs)

    transform=transforms.Compose([
    transforms.ToTensor(),
    transforms.Normalize((0.1307,), (0.3081,))
    ])

    testset = datasets.MNIST('./data', train=False,
                       transform=transform)

    test_loader = torch.utils.data.DataLoader(testset, **test_kwargs)

    return test_loader

def test(model, device, test_loader):
    model.eval()
    test_loss = 0
    correct_num = 0

    with torch.no_grad():
        num_correct_img = 1
        num_false_img = 1
        for idx, (data, target) in enumerate(test_loader):
            data, target = data.to(device), target.to(device)
            output = model(data)
            test_loss += F.cross_entropy(output, target, reduction='sum').item()
            predict = output.argmax(dim=1, keepdim=True)
            target = target.view_as(predict)
            correct_num += predict.eq(target).sum().item()
            for i, ((pred, targ), image_array) in enumerate(zip(zip(predict, target), data)):
                if pred == targ:
                    if random.randint(1,750) == 10:
                        fileName = 'pred_res/correct/c_' + str(num_correct_img) + '_targ_' + str(targ.item()) + '_pred' + str(pred.item()) + '_con_{:.4f}.png'.format(F.softmax(output,dim=1).cpu().numpy()[i, pred])
                        num_correct_img += 1
                        drawImage(fileName, image_array.reshape(28,28))
                else:
                    if random.randint(1,10) == 10:
                        fileName = 'pred_res/false/f_' + str(num_false_img) + '_targ_' + str(targ.item()) + '_pred' + str(pred.item()) + '_con_{:.4f}.png'.format(F.softmax(output,dim=1).cpu().numpy()[i, pred])
                        num_false_img += 1
                        drawImage(fileName, image_array.reshape(28,28))
    test_loss /= len(test_loader.dataset)
    accuracy = correct_num / len(test_loader.dataset)

    print('\nTest set: Average loss: {:.4f}, Accuracy: {}/{} ({:.4f}%)\n'.format(test_loss, correct_num, len(test_loader.dataset), 100.*accuracy))

    return 0

def fileCreate(fileName):
    if os.path.exists(fileName) is True:
        shutil.rmtree(fileName)
        os.makedirs(fileName)
    else:
        os.makedirs(fileName)

def drawImage(fileName, image_array):
    img = np.array(image_array.to("cpu"))
    img = 255 * (img * 0.3081 + 0.1307)
    img = img.astype(np.uint8)
    cv2.imwrite(fileName,img)

    return 0

if __name__ == '__main__':
    args = parse_args()
    main(args)