Refactor code to work with newer dataset

ndrplz · Oct 20, 2017 · 89e8d28 · 89e8d28
1 parent 3464dbf
commit 89e8d28
Show file tree

Hide file tree

Showing 2 changed files with 21 additions and 31 deletions.
diff --git a/capstone_traffic_light_classifier/traffic_light_dataset.py b/capstone_traffic_light_classifier/traffic_light_dataset.py
@@ -72,6 +72,21 @@ def load_batch(self, batch_size, augmentation=False):
 
         return X_batch, Y_batch
 
+    def print_statistics(self):
+        """
+        Print simple statistics on the number of samples in the dataset. 
+        :return: 
+        """
+        if not self.initialized:
+            raise IOError('Please initialize dataset first.')
+
+        color2label = {'none': 0, 'red': 1, 'yellow': 2, 'green': 3}
+
+        statistics = {}
+        for (color, num_label) in color2label.items():
+            statistics[color] = np.sum(self.dataset_npy[:, 1] == num_label)
+        print(statistics)
+
     @staticmethod
     def preprocess(x):
         """
@@ -113,26 +128,3 @@ def infer_label_from_frame_path(path):
         elif 'green' in path:
             label = 3
         return label
-
-    def print_statistics(self):
-
-        if not self.initialized:
-            raise IOError('Please initialize dataset first.')
-
-        color2label = {'none': 0, 'red': 1, 'yellow': 2, 'green': 3}
-
-        statistics = {}
-        for (color, num_label) in color2label.items():
-            statistics[color] = np.sum(self.dataset_npy[:, 1] == num_label)
-        print(statistics)
-
-if __name__ == '__main__':
-    # Init traffic light dataset
-    dataset = TrafficLightDataset()
-    # dataset_root = 'C:/Users/minotauro/Desktop/traffic_light_dataset'
-    # dataset.init_from_files(dataset_root, resize=(input_h, input_w))
-    # dataset.dump_to_npy('traffic_light_dataset.npy')
-    dataset.init_from_npy('traffic_light_dataset.npy')
-
-    # Load a batch of training data
-    x_batch, y_batch = dataset.load_batch(batch_size=16, augmentation=True)
diff --git a/capstone_traffic_light_classifier/train.py b/capstone_traffic_light_classifier/train.py
@@ -8,21 +8,19 @@
 if __name__ == '__main__':
 
     # Parameters
-    input_h, input_w = 64, 64    # Shape to which input is resized
+    input_h, input_w = 128, 128    # Shape to which input is resized
 
     # Init traffic light dataset
     dataset = TrafficLightDataset()
-    dataset_root = 'C:/Users/minotauro/Google Drive/SHARE/traffic_light_dataset'
-    dataset.init_from_files(dataset_root, resize=(input_h, input_w))
-    dataset.dump_to_npy('traffic_light_dataset.npy')
-    # dataset.init_from_npy('traffic_light_dataset.npy')
+    dataset_file = 'traffic_light_dataset_npy/traffic_light_dataset_mixed_resize_{}.npy'.format(input_h)
+    dataset.init_from_npy(dataset_file)
 
     # Define model
-    classifier = TrafficLightClassifier(input_shape=[input_h, input_w], learning_rate=1e-4)
+    classifier = TrafficLightClassifier(input_shape=[input_h, input_w], learning_rate=1e-4, verbose=True)
 
     # Checkpoint stuff
-    saver = tf.train.Saver()         # saver to save the model after each epoch
-    checkpoint_dir = './checkpoint_2'  # checkpoint directory
+    saver = tf.train.Saver()  # saver to save the model after each epoch
+    checkpoint_dir = './checkpoint_mixed_{}'.format(input_h)  # checkpoint directory
     if not exists(checkpoint_dir):
         makedirs(checkpoint_dir)