AttributeError: Layer vgg16 has multiple inbound nodes, hence the notion of "layer input" is ill-defined. Use get_input_at(node_index) instead.

[gaody9527]

when I run the code ,the following question appeared
AttributeError: Layer vgg16 has multiple inbound nodes, hence the notion of "layer input" is ill-defined. Use get_input_at(node_index) instead.
i change it as
gradient_function = K.function([model.layers[0].get_input_at(1)], [conv_output, grads])
but it also does not work
i do not understand what the function want? any help will be appreciated.

[kobotschick]

My solution was not to define the model as sequential, but to add the target layer directly.
x = Lambda(target_layer, output_shape = target_category_loss_output_shape)(input_model.output)
model = Model(inputs=input_model.input, outputs=x)

The grad_cam function looks like this now:

    nb_classes = 1000
    target_layer = lambda x: target_category_loss(x, category_index, nb_classes)
    x = Lambda(target_layer, output_shape = target_category_loss_output_shape)(input_model.output)
    model = Model(inputs=input_model.input, outputs=x)
    model.summary()
    loss = K.sum(model.output)
    conv_output =  [l for l in model.layers if l.name is layer_name][0].output
    grads = normalize(_compute_gradients(loss, [conv_output])[0])
    gradient_function = K.function([model.input], [conv_output, grads])

    output, grads_val = gradient_function([image])
    output, grads_val = output[0, :], grads_val[0, :, :, :]

    weights = np.mean(grads_val, axis = (0, 1))
    cam = np.ones(output.shape[0 : 2], dtype = np.float32)

    for i, w in enumerate(weights):
        cam += w * output[:, :, i]

    cam = cv2.resize(cam, (224, 224))
    cam = np.maximum(cam, 0)
    heatmap = cam / np.max(cam)

    #Return to BGR [0..255] from the preprocessed image
    image = image[0, :]
    image -= np.min(image)
    image = np.minimum(image, 255)

    cam = cv2.applyColorMap(np.uint8(255*heatmap), cv2.COLORMAP_JET)
    cam = np.float32(cam) + np.float32(image)
    cam = 255 * cam / np.max(cam)
    return np.uint8(cam), heatmap´´´

[gaody9527]

@kobotschick Thanks you for your proposals， I have dealed with the problem to follow the https://github.com/vense/keras-grad-cam/blob/master/grad-cam.py

[jurgyy]

@kobotschick One problem I had with your solution was a message that a list doesn't have an attributed called dtype. The solution was to change gradient_function = K.function([model.input], [conv_output, grads]) to gradient_function = K.function(model.input, [conv_output, grads])

[jayanthsrinivas]

@kobotschick Where is Model() function defined?

[somdipdey]

full grad_cam function along with import as follows:

from keras.models import Model

def grad_cam(input_model, image, category_index, layer_name):
    nb_classes = 1000
    target_layer = lambda x: target_category_loss(x, category_index, nb_classes)
    x = Lambda(target_layer, output_shape = target_category_loss_output_shape)(input_model.output)
    model = Model(inputs=input_model.input, outputs=x)
    model.summary()
    loss = K.sum(model.output)
    conv_output =  [l for l in model.layers if l.name is layer_name][0].output
    grads = normalize(_compute_gradients(loss, [conv_output])[0])
    gradient_function = K.function([model.input], [conv_output, grads])

    output, grads_val = gradient_function([image])
    output, grads_val = output[0, :], grads_val[0, :, :, :]

    weights = np.mean(grads_val, axis = (0, 1))
    cam = np.ones(output.shape[0 : 2], dtype = np.float32)

    for i, w in enumerate(weights):
        cam += w * output[:, :, i]

    cam = cv2.resize(cam, (224, 224))
    cam = np.maximum(cam, 0)
    heatmap = cam / np.max(cam)

    #Return to BGR [0..255] from the preprocessed image
    image = image[0, :]
    image -= np.min(image)
    image = np.minimum(image, 255)

    cam = cv2.applyColorMap(np.uint8(255*heatmap), cv2.COLORMAP_JET)
    cam = np.float32(cam) + np.float32(image)
    cam = 255 * cam / np.max(cam)
    return np.uint8(cam), heatmap

[PowerOfCreation]

For everyone who is facing the same issue: I forked this repository and fixed all errors I was facing.

#27
https://github.com/PowerOfCreation/keras-grad-cam

[jple]

The following change works for me :
gradient_function = K.function([model.layers[0].get_input_at(1)], [conv_output, grads])
instead of
gradient_function = K.function([model.layers[0].get_input_at(0)], [conv_output, grads])