Merge remote-tracking branch 'origin/master'

# Conflicts:
#	createNeuralNetworkDatabase.py

createNeuralNetworkDatabase.py

@@ -3,37 +3,46 @@ import cv2
 import matplotlib
 import numpy as np
 import matplotlib.pyplot as plt
+import pandas as pd
+from matplotlib.pyplot import imshow
 
 path_potatoes = 'neural_network\\images\\potatoes'
 path_beetroot = 'neural_network\\images\\beetroot'
-size = 250
+size = 100
 
 #POTATOES
-training_data_potatoes = []
+image_data = []
+label_data = []
 for img in os.listdir(path_potatoes):
     pic = cv2.imread(os.path.join(path_potatoes,img))
     pic = cv2.cvtColor(pic,cv2.COLOR_BGR2RGB)
     pic = cv2.resize(pic,(size,size))
-    training_data_potatoes.append([pic])
+    image_data.append([pic])
+    label_data.append(1)
 
+#np.save(os.path.join('neural_network','potatoes-dataset'),np.array(training_data_potatoes))
 
-np.save(os.path.join('neural_network','potatoes-dataset'),np.array(training_data_potatoes))
-
-saved_potatoes = np.load(os.path.join('neural_network','potatoes-dataset.npy'))
+#saved_potatoes = np.load(os.path.join('neural_network','potatoes-dataset.npy'))
 
 #BEETROOT
-training_data_beetroot = []
 for img in os.listdir(path_beetroot):
     pic = cv2.imread(os.path.join(path_beetroot,img))
     pic = cv2.cvtColor(pic,cv2.COLOR_BGR2RGB)
     pic = cv2.resize(pic,(size,size))
-    training_data_beetroot.append([pic])
+    image_data.append([pic])
+    label_data.append(0)
 
-np.save(os.path.join('neural_network','beetroot-dataset'),np.array(training_data_beetroot))
+#np.save(os.path.join('neural_network','beetroot-dataset'),np.array(training_data_beetroot))
 
-saved_potatoes = np.load(os.path.join('neural_network','beetroot-dataset.npy'))
+#saved_potatoes = np.load(os.path.join('neural_network','beetroot-dataset.npy'))
 
-dict = {
-    'beetroots': training_data_beetroot,
-    'potatoes': training_data_potatoes
-}
+np.save(os.path.join('neural_network','image-dataset'),np.array(image_data))
+np.save(os.path.join('neural_network','label-dataset'),np.array(label_data))
+
+saved_images = np.load(os.path.join('neural_network','image-dataset.npy'))
+
+print(saved_images.shape)
+
+plt.imshow(saved_images[0].reshape(size,size,3))
+plt.imshow(np.array(image_data[0]).reshape(size,size,3))
+plt.show()

neuralNetwork.py

@@ -0,0 +1,54 @@
+import numpy as np
+import torch
+import torch.nn as nn
+import torch.optim as optim
+from matplotlib.pyplot import imshow
+import numpy as np
+from matplotlib.pyplot import imshow
+import matplotlib.pyplot as ppl
+
+def plotdigit(image):
+    img = np.reshape(image, (-250, 250))
+    imshow(img, cmap='Greys', vmin=0, vmax=255)
+    ppl.show()
+
+train_images = np.load('neural_network\\image-dataset.npy')
+print(train_images.shape)
+
+train_labels = np.load('neural_network\\label-dataset.npy')
+
+train_images = train_images / 255
+train_labels = train_labels / 255
+
+train_images = [torch.tensor(image, dtype=torch.float32) for image in train_images]
+print(train_images[0].shape)
+
+train_labels = [torch.tensor(label, dtype=torch.long) for label in train_labels]
+
+
+input_dim = 100*100*3
+output_dim = 2
+
+model = nn.Sequential(
+    nn.Linear(input_dim, output_dim),
+    nn.LogSoftmax()
+)
+
+def train(model, n_iter):
+
+    criterion = nn.NLLLoss()
+    optimizer = optim.SGD(model.parameters(), lr=0.001)
+
+    for epoch in range(n_iter):
+        for image, label in zip(train_images, train_labels):
+            optimizer.zero_grad()
+
+            output = model(image)
+            loss = criterion(output.unsqueeze(0), label.unsqueeze(0))
+            loss.backward()
+            optimizer.step()
+
+        print(f'epoch: {epoch:03}')
+
+
+train(model, 100)