python - 如何在训练有素的网络上实现 Grad-CAM

Question

我已经训练了一个网络，并以 mynetwork.model 的形式保存了它。我想使用我自己的模型而不是 VGG16 或 ResNet 等应用 gradcam。
apply_gradcam.py

# import the necessary packages
from Grad_CAM.gradcam import GradCAM
from tensorflow.keras.applications import ResNet50
from tensorflow.keras.applications import VGG16
from tensorflow.keras.preprocessing.image import img_to_array
from tensorflow.keras.preprocessing.image import load_img
from tensorflow.keras.applications import imagenet_utils
from tensorflow.keras.models import load_model
import numpy as np
import argparse
import imutils
import cv2


# construct the argument parser and parse the arguments
ap = argparse.ArgumentParser()
ap.add_argument("-i", "--image", required=True,
    help="path to the input image")
ap.add_argument("-m", "--model", type=str, default="vgg",
    #choices=("vgg", "resnet"),
    help="model to be used")
args = vars(ap.parse_args())


# initialize the model to be VGG16
Model = VGG16
# check to see if we are using ResNet
if args["model"] == "resnet":
    Model = ResNet50
# load the pre-trained CNN from disk
print("[INFO] loading model...")
model = Model(weights="imagenet")

# load the original image from disk (in OpenCV format) and then
# resize the image to its target dimensions
orig = cv2.imread(args["image"])
resized = cv2.resize(orig, (224, 224))
# load the input image from disk (in Keras/TensorFlow format) and
# preprocess it
image = load_img(args["image"], target_size=(224, 224))
image = img_to_array(image)
image = np.expand_dims(image, axis=0)
image = imagenet_utils.preprocess_input(image)

# use the network to make predictions on the input image and find
# the class label index with the largest corresponding probability
preds = model.predict(image)
i = np.argmax(preds[0])
# decode the ImageNet predictions to obtain the human-readable label
decoded = imagenet_utils.decode_predictions(preds)
(imagenetID, label, prob) = decoded[0][0]
label = "{}: {:.2f}%".format(label, prob * 100)
print("[INFO] {}".format(label))

# initialize our gradient class activation map and build the heatmap
cam = GradCAM(model, i)
heatmap = cam.compute_heatmap(image)
# resize the resulting heatmap to the original input image dimensions
# and then overlay heatmap on top of the image
heatmap = cv2.resize(heatmap, (orig.shape[1], orig.shape[0]))
(heatmap, output) = cam.overlay_heatmap(heatmap, orig, alpha=0.5)

cv2.rectangle(output, (0, 0), (340, 40), (0, 0, 0), -1)
cv2.putText(output, label, (10, 25), cv2.FONT_HERSHEY_SIMPLEX,
    0.8, (255, 255, 255), 2)
# display the original image and resulting heatmap and output image
# to our screen
output = np.vstack([orig, heatmap, output])
output = imutils.resize(output, height=700)
cv2.imshow("Output", output)
cv2.waitKey(0)

gradcam.py

from tensorflow.keras.models import Model
import tensorflow as tf
import numpy as np
import cv2


class GradCAM:
    def __init__(self, model, classIdx, layerName=None):
        # store the model, the class index used to measure the class
        # activation map, and the layer to be used when visualizing
        # the class activation map
        self.model = model
        self.classIdx = classIdx
        self.layerName = layerName
        # if the layer name is None, attempt to automatically find
        # the target output layer
        if self.layerName is None:
            self.layerName = self.find_target_layer()


    def find_target_layer(self):
        # attempt to find the final convolutional layer in the network
        # by looping over the layers of the network in reverse order
        for layer in reversed(self.model.layers):
            # check to see if the layer has a 4D output
            if len(layer.output_shape) == 4:
                return layer.name
        # otherwise, we could not find a 4D layer so the GradCAM
        # algorithm cannot be applied
        raise ValueError("Could not find 4D layer. Cannot apply GradCAM.")


    def compute_heatmap(self, image, eps=1e-8):
        # construct our gradient model by supplying (1) the inputs
        # to our pre-trained model, (2) the output of the (presumably)
        # final 4D layer in the network, and (3) the output of the
        # softmax activations from the model
        gradModel = Model(
            inputs=[self.model.inputs],
            outputs=[self.model.get_layer(self.layerName).output,
                     self.model.output])

        # record operations for automatic differentiation
        with tf.GradientTape() as tape:
            # cast the image tensor to a float-32 data type, pass the
            # image through the gradient model, and grab the loss
            # associated with the specific class index
            inputs = tf.cast(image, tf.float32)
            (convOutputs, predictions) = gradModel(inputs)
            loss = predictions[:, self.classIdx]
        # use automatic differentiation to compute the gradients
        grads = tape.gradient(loss, convOutputs)

        # compute the guided gradients
        castConvOutputs = tf.cast(convOutputs > 0, "float32")
        castGrads = tf.cast(grads > 0, "float32")
        guidedGrads = castConvOutputs * castGrads * grads
        # the convolution and guided gradients have a batch dimension
        # (which we don't need) so let's grab the volume itself and
        # discard the batch
        convOutputs = convOutputs[0]
        guidedGrads = guidedGrads[0]

        # compute the average of the gradient values, and using them
        # as weights, compute the ponderation of the filters with
        # respect to the weights
        weights = tf.reduce_mean(guidedGrads, axis=(0, 1))
        cam = tf.reduce_sum(tf.multiply(weights, convOutputs), axis=-1)

        # grab the spatial dimensions of the input image and resize
        # the output class activation map to match the input image
        # dimensions
        (w, h) = (image.shape[2], image.shape[1])
        heatmap = cv2.resize(cam.numpy(), (w, h))
        # normalize the heatmap such that all values lie in the range
        # [0, 1], scale the resulting values to the range [0, 255],
        # and then convert to an unsigned 8-bit integer
        numer = heatmap - np.min(heatmap)
        denom = (heatmap.max() - heatmap.min()) + eps
        heatmap = numer / denom
        heatmap = (heatmap * 255).astype("uint8")
        # return the resulting heatmap to the calling function
        return heatmap

    def overlay_heatmap(self, heatmap, image, alpha=0.5,
                        colormap=cv2.COLORMAP_VIRIDIS):
        # apply the supplied color map to the heatmap and then
        # overlay the heatmap on the input image
        heatmap = cv2.applyColorMap(heatmap, colormap)
        output = cv2.addWeighted(image, alpha, heatmap, 1 - alpha, 0)
        # return a 2-tuple of the color mapped heatmap and the output,
        # overlaid image
        return (heatmap, output)

正如您在 apply_gradcam.py 中看到的，使用了 VGG16 或 ResNet 预训练模型。我想使用我自己训练过的模型来执行 gradcam。出于这个原因，我评论了以下几行:

   # initialize the model to be VGG16
    Model = VGG16
    # check to see if we are using ResNet
    if args["model"] == "resnet":
        Model = ResNet50
    # load the pre-trained CNN from disk
    print("[INFO] loading model...")
    model = Model(weights="imagenet")

我用过

model = load_model(args["model"]) 

为了使用我自己的模型。然后我执行了:

 python apply_gradcam.py --image /home/antonis/IM0001.jpeg --model /home/antonis/mynetwork.model

但是，我收到以下错误:

ValueError: `decode_predictions` expects a batch of predictions (i.e.
a 2D array of shape (samples, 1000)). Found array with shape: (1, 3)

这是预期的，因为模型输出 ImageNet 类(1000 维)，而我的模型返回超过 2 个类的预测。
我想知道如何解决这个问题并使用我自己的模型应用 gradcam。

Answer 1

我不明白的一件事是，如果您有自己的分类器( 2 )，为什么还要使用 imagenet_utils.decode_predictions ?我不确定我下面的回答是否会让你满意。但这里有一些指针。
数据集

import tensorflow as tf
import numpy as np 

(x_train, y_train), (x_test, y_test) = tf.keras.datasets.cifar10.load_data()

# train set / data 
x_train = x_train.astype('float32') / 255
# train set / target 
y_train = tf.keras.utils.to_categorical(y_train , num_classes=10)

# validation set / data 
x_test = x_test.astype('float32') / 255
# validation set / target 
y_test = tf.keras.utils.to_categorical(y_test, num_classes=10)

print(x_train.shape, y_train.shape)
print(x_test.shape, y_test.shape)  
# (50000, 32, 32, 3) (50000, 10)
# (10000, 32, 32, 3) (10000, 10

型号

input = tf.keras.Input(shape=(32,32,3))
efnet = tf.keras.applications.EfficientNetB0(weights='imagenet',
                                             include_top = False, 
                                             input_tensor = input)
# Now that we apply global max pooling.
gap = tf.keras.layers.GlobalMaxPooling2D()(efnet.output)

# Finally, we add a classification layer.
output = tf.keras.layers.Dense(10, activation='softmax')(gap)

# bind all
func_model = tf.keras.Model(efnet.input, output)

编译运行

func_model.compile(
          loss      = tf.keras.losses.CategoricalCrossentropy(),
          metrics   = tf.keras.metrics.CategoricalAccuracy(),
          optimizer = tf.keras.optimizers.Adam())
# fit 
func_model.fit(x_train, y_train, batch_size=128, epochs=15, verbose = 2)

Epoch 14/15
391/391 - 13s - loss: 0.1479 - categorical_accuracy: 0.9491
Epoch 15/15
391/391 - 13s - loss: 0.1505 - categorical_accuracy: 0.9481

毕业凸轮
和你的设置一样。

from tensorflow.keras.models import Model
import tensorflow as tf
import numpy as np
import cv2

class GradCAM:
    def __init__(self, model, classIdx, layerName=None):
        # store the model, the class index used to measure the class
        # activation map, and the layer to be used when visualizing
        # the class activation map
        self.model = model
        self.classIdx = classIdx
        self.layerName = layerName
        # if the layer name is None, attempt to automatically find
        # the target output layer
        if self.layerName is None:
            self.layerName = self.find_target_layer()

    def find_target_layer(self):
        # attempt to find the final convolutional layer in the network
        # by looping over the layers of the network in reverse order
        for layer in reversed(self.model.layers):
            # check to see if the layer has a 4D output
            if len(layer.output_shape) == 4:
                return layer.name
        # otherwise, we could not find a 4D layer so the GradCAM
        # algorithm cannot be applied
        raise ValueError("Could not find 4D layer. Cannot apply GradCAM.")


    def compute_heatmap(self, image, eps=1e-8):
        # construct our gradient model by supplying (1) the inputs
        # to our pre-trained model, (2) the output of the (presumably)
        # final 4D layer in the network, and (3) the output of the
        # softmax activations from the model
        gradModel = Model(
            inputs=[self.model.inputs],
            outputs=[self.model.get_layer(self.layerName).output, self.model.output])

        # record operations for automatic differentiation
        with tf.GradientTape() as tape:
            # cast the image tensor to a float-32 data type, pass the
            # image through the gradient model, and grab the loss
            # associated with the specific class index
            inputs = tf.cast(image, tf.float32)
            (convOutputs, predictions) = gradModel(inputs)
            
            loss = predictions[:, tf.argmax(predictions[0])]
    
        # use automatic differentiation to compute the gradients
        grads = tape.gradient(loss, convOutputs)

        # compute the guided gradients
        castConvOutputs = tf.cast(convOutputs > 0, "float32")
        castGrads = tf.cast(grads > 0, "float32")
        guidedGrads = castConvOutputs * castGrads * grads
        # the convolution and guided gradients have a batch dimension
        # (which we don't need) so let's grab the volume itself and
        # discard the batch
        convOutputs = convOutputs[0]
        guidedGrads = guidedGrads[0]

        # compute the average of the gradient values, and using them
        # as weights, compute the ponderation of the filters with
        # respect to the weights
        weights = tf.reduce_mean(guidedGrads, axis=(0, 1))
        cam = tf.reduce_sum(tf.multiply(weights, convOutputs), axis=-1)

        # grab the spatial dimensions of the input image and resize
        # the output class activation map to match the input image
        # dimensions
        (w, h) = (image.shape[2], image.shape[1])
        heatmap = cv2.resize(cam.numpy(), (w, h))
        # normalize the heatmap such that all values lie in the range
        # [0, 1], scale the resulting values to the range [0, 255],
        # and then convert to an unsigned 8-bit integer
        numer = heatmap - np.min(heatmap)
        denom = (heatmap.max() - heatmap.min()) + eps
        heatmap = numer / denom
        heatmap = (heatmap * 255).astype("uint8")
        # return the resulting heatmap to the calling function
        return heatmap

    def overlay_heatmap(self, heatmap, image, alpha=0.5,
                        colormap=cv2.COLORMAP_VIRIDIS):
        # apply the supplied color map to the heatmap and then
        # overlay the heatmap on the input image
        heatmap = cv2.applyColorMap(heatmap, colormap)
        output = cv2.addWeighted(image, alpha, heatmap, 1 - alpha, 0)
        # return a 2-tuple of the color mapped heatmap and the output,
        # overlaid image
        return (heatmap, output)

预测

image = cv2.imread('/content/dog.jpg')
image = cv2.resize(image, (32, 32))
image = image.astype('float32') / 255
image = np.expand_dims(image, axis=0)

preds = func_model.predict(image) 
i = np.argmax(preds[0])

获取模型的图层名称

for idx in range(len(func_model.layers)):
  print(func_model.get_layer(index = idx).name)

# we picked `block5c_project_con` layer 

传递到 GradCAM类(class)

icam = GradCAM(func_model, i, 'block5c_project_conv') 
heatmap = icam.compute_heatmap(image)
heatmap = cv2.resize(heatmap, (32, 32))

image = cv2.imread('/content/dog.jpg')
image = cv2.resize(image, (32, 32))
print(heatmap.shape, image.shape)

(heatmap, output) = icam.overlay_heatmap(heatmap, image, alpha=0.5)

可视化

fig, ax = plt.subplots(1, 3)

ax[0].imshow(heatmap)
ax[1].imshow(image)
ax[2].imshow(output)

引用 Grad-CAM class activation visualization