Problem with training U-Net while using ImageDataGenerator on multiclass segmentation

Question

The task I am dealing with is multiclass segmentation (0-3 classes on each image). I had a working U-Net model and could train on small dataset just fine, then I augmented the dataset and now I have almost 15k 512x512 grayscale images. I've naturally encountered an issue with not having enough HW resources (RAM, GPU), so I decided to switch to google colab and to go with ImageDataGenerator. I've encountered this issue not being able to solve so far.

InvalidArgumentError: Conv2DSlowBackpropInput: Size of out_backprop doesn't match computed: actual = 16, computed = 32 spatial_dim: 2 input: 64 filter: 2 output: 16 stride: 2 dilation: 1 [[node model/conv2d_transpose_1/conv2d_transpose (defined at /usr/local/lib/python3.7/dist-packages/keras/backend.py:5360) ]] [Op:__inference_train_function_3151]

The only explanation for me is I am not using the generator well. I've structured the data as:

path_to_dataset
│
└───images_dir
│   │
│   └─── images_subdir
│       │   img1.png
│       │   img2.png
│       │   ...
│   
└───masks_dir
│   │
│   └─── masks_subdir
│       │   img1.png
│       │   img2.png
│       │   ...

The subdirectories are there to only make ImageDataGenerator work.

data_gen_args = dict(rescale=1./255,)
image_datagen = ImageDataGenerator(**data_gen_args)
mask_datagen = ImageDataGenerator(**data_gen_args)
# image_datagen.fit(images)
# mask_datagen.fit(masks)
# Provide the same seed and keyword arguments to the fit and flow methods
seed = 1
image_generator = image_datagen.flow_from_directory(
    '/content/drive/MyDrive/DP/preprocess_images/images/final_ds/orig_folder/',
    batch_size=16,
    class_mode=None,
    # color_mode='grayscale',
    seed=seed)
mask_generator = mask_datagen.flow_from_directory(
    '/content/drive/MyDrive/DP/preprocess_images/images/final_ds/seg_greyscale_folder/',
    batch_size=16,
    class_mode=None,
    # color_mode='grayscale',
    seed=seed)
# combine generators into one which yields image and masks
train_generator = zip(image_generator, mask_generator)

callbacks = [
    ModelCheckpoint('unet_512.h5', verbose=1, save_best_only=True),
    EarlyStopping(patience=5, monitor='val_loss'),
    TensorBoard(log_dir='logs_unet512')
]

history = model.fit(train_generator,
                    verbose=1,
                    epochs=50,
                    callbacks=callbacks,
                    # class_weight=class_weights,
                    shuffle=False)

I didn't deal with creating data generator for validation data so far, since I'm not even able to make this part work.

And for the curious ones, here's the model.

    # IMG_HEIGHT=512, IMG_WIDTH=512, IMG_CHANNELS=1
    inputs = Input((IMG_HEIGHT, IMG_WIDTH, IMG_CHANNELS))
    s = inputs

    # Contraction path
    c1 = Conv2D(16, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(s)
    c1 = Dropout(0.1)(c1)
    c1 = Conv2D(16, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(c1)
    p1 = MaxPooling2D((2, 2))(c1)

    c2 = Conv2D(32, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(p1)
    c2 = Dropout(0.1)(c2)
    c2 = Conv2D(32, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(c2)
    p2 = MaxPooling2D((2, 2))(c2)

    c3 = Conv2D(64, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(p2)
    c3 = Dropout(0.2)(c3)
    c3 = Conv2D(64, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(c3)
    p3 = MaxPooling2D((2, 2))(c3)

    c4 = Conv2D(128, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(p3)
    c4 = Dropout(0.2)(c4)
    c4 = Conv2D(128, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(c4)
    p4 = MaxPooling2D(pool_size=(2, 2))(c4)

    c5 = Conv2D(256, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(p4)
    c5 = Dropout(0.3)(c5)
    c5 = Conv2D(256, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(c5)

    # Expansive path
    u6 = Conv2DTranspose(128, (2, 2), strides=(2, 2), padding='same')(c5)
    u6 = concatenate([u6, c4])
    c6 = Conv2D(128, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(u6)
    c6 = Dropout(0.2)(c6)
    c6 = Conv2D(128, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(c6)

    u7 = Conv2DTranspose(64, (2, 2), strides=(2, 2), padding='same')(c6)
    u7 = concatenate([u7, c3])
    c7 = Conv2D(64, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(u7)
    c7 = Dropout(0.2)(c7)
    c7 = Conv2D(64, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(c7)

    u8 = Conv2DTranspose(32, (2, 2), strides=(2, 2), padding='same')(c7)
    u8 = concatenate([u8, c2])
    c8 = Conv2D(32, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(u8)
    c8 = Dropout(0.1)(c8)
    c8 = Conv2D(32, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(c8)

    u9 = Conv2DTranspose(16, (2, 2), strides=(2, 2), padding='same')(c8)
    u9 = concatenate([u9, c1], axis=3)
    c9 = Conv2D(16, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(u9)
    c9 = Dropout(0.1)(c9)
    c9 = Conv2D(16, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(c9)

    # n_classes=4
    outputs = Conv2D(n_classes, (1, 1), activation='softmax')(c9)

    model = Model(inputs=[inputs], outputs=[outputs])

Edit: also planning to get the number of filters up, so far I'm running the model that worked before on my personal laptop

Problem with training U-Net while using ImageDataGenerator on multiclass segmentation

Answers (1)

Related Questions