Q&A
CNNを学習させ、胸部X線の疾患を検出するプログラムを書いています。
現在は練習として少ない画像で練習しています。そのため、細菌に感染した肺とウイルスに感染した肺の画像で判別を行っております。
自分のGoogle driveに保存してある画像をリサイズし、CNNに読み込ませています。
CNNのモデルを学習させていましたが、下記のエラーが出てきます。
ValueError: in user code: File "/usr/local/lib/python3.7/dist-packages/keras/engine/training.py", line 878, in train_function * return step_function(self, iterator) File "/usr/local/lib/python3.7/dist-packages/keras/engine/training.py", line 867, in step_function ** outputs = model.distribute_strategy.run(run_step, args=(data,)) File "/usr/local/lib/python3.7/dist-packages/keras/engine/training.py", line 860, in run_step ** outputs = model.train_step(data) File "/usr/local/lib/python3.7/dist-packages/keras/engine/training.py", line 810, in train_step y, y_pred, sample_weight, regularization_losses=self.losses) File "/usr/local/lib/python3.7/dist-packages/keras/engine/compile_utils.py", line 201, in __call__ loss_value = loss_obj(y_t, y_p, sample_weight=sw) File "/usr/local/lib/python3.7/dist-packages/keras/losses.py", line 141, in __call__ losses = call_fn(y_true, y_pred) File "/usr/local/lib/python3.7/dist-packages/keras/losses.py", line 245, in call ** return ag_fn(y_true, y_pred, **self._fn_kwargs) File "/usr/local/lib/python3.7/dist-packages/keras/losses.py", line 1665, in categorical_crossentropy y_true, y_pred, from_logits=from_logits, axis=axis) File "/usr/local/lib/python3.7/dist-packages/keras/backend.py", line 4994, in categorical_crossentropy target.shape.assert_is_compatible_with(output.shape) ValueError: Shapes (None, 1) and (None, 2) are incompatible
ソースコードはこちらになります。
Python
1from keras.utils import np_utils 2from keras.models import Sequential 3from keras.layers.convolutional import MaxPooling2D 4from keras.layers import Activation, Conv2D, Flatten, Dense, Dropout 5from keras.engine import base_layer_utils 6from sklearn.model_selection import train_test_split 7import numpy as np 8import keras 9import glob 10import os 11import cv2 12!pip install tensorflow 13!pip install tensorflow-gpu 14!pip install keras 15 16num_classes = 2 17 18bacteria = [] 19virus = [] 20 21files = glob.glob("/content/drive/MyDrive/resize/bacteria_small/*.*") 22i = 1 23for file in files: 24 img_bacteria = cv2.imread(file) 25 img_bacteria = np.array(img_bacteria) 26 bacteria.append(img_bacteria) 27bacteria = np.array(bacteria) 28 29files = glob.glob("/content/drive/MyDrive/resize/virus_small/*.*") 30i = 1 31for file in files: 32 img_virus = cv2.imread(file) 33 img_virus = np.array(img_virus) 34 virus.append(img_virus) 35virus = np.array(virus) 36 37np.savez('gazou_np', bacteria, virus) 38 39a = np.load('/content/gazou_np.npz') 40 41x = np.concatenate([a['arr_0'],a['arr_1']]) 42bacteria_label = np.array([0]*len(a['arr_0'])) 43virus_label = np.array([1]*len(a['arr_1'])) 44y = np.concatenate([bacteria_label, virus_label]) 45y = np.array(y) 46 47x_train, x_test, y_train, y_test = train_test_split(x, y, test_size = 0.20) 48 49 50x_train = x_train.astype('float32')/255 51x_test = x_test.astype('float32')/255 52 53input_shape = x_train.shape 54 55model = Sequential() 56 57model.add(Conv2D(32, (3, 3), padding='same', input_shape = (input_shape))) 58model.add(Activation('relu')) 59model.add(MaxPooling2D(pool_size=(2, 2))) 60model.add(Dropout(0.25)) 61 62model.add(Conv2D(64, (3, 3), padding='same')) 63model.add(Activation('relu')) 64model.add(MaxPooling2D(pool_size=(2, 2))) 65model.add(Dropout(0.25)) 66 67model.add(Conv2D(128, (3, 3), padding='same')) 68model.add(Activation('relu')) 69model.add(MaxPooling2D(pool_size=(2, 2))) 70model.add(Dropout(0.25)) 71 72model.add(Flatten()) 73model.add(Dense1(16)) 74model.add(Activation('relu')) 75model.add(Dense2(64)) 76model.add(Activation('relu')) 77model.add(Dense3(128)) 78model.add(Activation('relu')) 79model.add(Dropout(0.5)) 80model.add(Dense(num_classes, activation='softmax')) 81 82model.summary() 83 84model.compile(loss='categorical_crossentropy', 85 optimizer='Adam', 86 metrics=['accuracy']) 87 88 89history = model.fit(x_train, y_train, epochs=50, validation_data=(x_test,y_test))
model.summary()の実行結果です。
Model: "sequential_1" _________________________________________________________________ Layer (type) Output Shape Param # ================================================================= conv2d_3 (Conv2D) (None, 64, 64, 32) 896 activation_3 (Activation) (None, 64, 64, 32) 0 max_pooling2d_3 (MaxPooling (None, 32, 32, 32) 0 2D) dropout_3 (Dropout) (None, 32, 32, 32) 0 conv2d_4 (Conv2D) (None, 32, 32, 64) 18496 activation_4 (Activation) (None, 32, 32, 64) 0 max_pooling2d_4 (MaxPooling (None, 16, 16, 64) 0 2D) dropout_4 (Dropout) (None, 16, 16, 64) 0 conv2d_5 (Conv2D) (None, 16, 16, 128) 73856 activation_5 (Activation) (None, 16, 16, 128) 0 max_pooling2d_5 (MaxPooling (None, 8, 8, 128) 0 2D) dropout_5 (Dropout) (None, 8, 8, 128) 0 flatten_1 (Flatten) (None, 8192) 0 dense (Dense) (None, 16) 131088 activation_6 (Activation) (None, 16) 0 dense_1 (Dense) (None, 64) 1088 activation_7 (Activation) (None, 64) 0 dense_2 (Dense) (None, 128) 8320 activation_8 (Activation) (None, 128) 0 dropout_6 (Dropout) (None, 128) 0 dense_3 (Dense) (None, 2) 258 ================================================================= Total params: 234,002 Trainable params: 234,002 Non-trainable params: 0 _________________________________________________________________
なぜエラーが出てくるのかがわかりません。どなたか解決方法を教えてください。
よろしくお願いします。
回答1件
0
ベストアンサー
出力層の期待するラベルの形状 (バッチサイズ, 2)
実際に与えられたラベルの形状 (バッチサイズ, 1)
で一致していないというエラーです。
ラベルを Ont-Hot 表現に変換してから、モデルに与えてください。
Keras だと to_categorical() を使うと変換できます。
Numpyユーティリティ - Keras Documentation
例: [A, A, B, B, B] -> [[1, 0], [1, 0], [0, 1], [0, 1], [0, 1]]
投稿2021/12/09 09:37
編集2021/12/09 09:38
総合スコア21956
あなたの回答
tips
プレビュー
バッドをするには、ログインかつ
こちらの条件を満たす必要があります。
2021/12/09 10:12