错误的输入形状densenet212

Question

我正在尝试在DenseNet121中为灰度图像创建梯度凸轮显着图。我在运行我的代码时遇到了问题，无法使我的图像符合所需的输入形状。

这是我的热图代码

def get_class_activation_map(path) :
    
    img_path =  path 
    img = cv2.imread(img_path)
    img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
    img = cv2.resize(img, (150, 150))
    img = np.expand_dims(img,axis=0)
    
    
    predict = model.predict(img)
    target_class = np.argmax(predict[0])
    last_conv = model.get_layer('conv2d_119')
    grads =K.gradients(model.output[:,target_class],last_conv.output)[0]
    pooled_grads = K.mean(grads,axis=(0,1,2))
    iterate = K.function([model.input],[pooled_grads,last_conv.output[0]])
    pooled_grads_value,conv_layer_output = iterate([img])
    
    for i in range(512):
        conv_layer_output[:,:,i] *= pooled_grads_value[i]
    
    heatmap = np.mean(conv_layer_output,axis=-1)
    
    for x in range(heatmap.shape[0]):
        for y in range(heatmap.shape[1]):
            heatmap[x,y] = np.max(heatmap[x,y],0)
    heatmap = np.maximum(heatmap,0)
    heatmap /= np.max(heatmap)
    plt.imshow(heatmap)

我得到的错误是

ValueError: Error when checking input: expected densenet121_input to have shape (150, 150, 1) but got array with shape (1, 150, 150)

你能告诉我如何改变我的代码，使它改变尺寸本身吗？

Answer 1

在执行直线预测之前：

predict = model.predict(img)

执行以下操作：

img = np.moveaxis(img, -1, 0)

这将使形状从channels_first反转为channels_last，这是您的模型中所期望的。