无法为多输出添加标准指标 model

Question

I have classification + detection model of cats and dogs based on MobileNet v2. It trains well, but now I want to add metrics for it and I can't do that. Here is the main part of code:

def localization_loss(y_true, yhat):            
    delta_coord = tf.reduce_sum(tf.square(y_true[:,:2] - yhat[:,:2]))          
    h_true = y_true[:,3] - y_true[:,1] 
    w_true = y_true[:,2] - y_true[:,0]
    h_pred = yhat[:,3] - yhat[:,1] 
    w_pred = yhat[:,2] - yhat[:,0] 
    delta_size = tf.reduce_sum(tf.square(w_true - w_pred) + tf.square(h_true-h_pred))
    return delta_coord + delta_size

classloss = tf.keras.losses.BinaryCrossentropy()
regressloss = localization_loss

opt = tf.keras.optimizers.Adam(learning_rate=0.0001, decay=0.001)

model.compile(
    optimizer = opt,
    loss=[classloss, regressloss],
    # metrics=["accuracy", "meaniou"],
)
hist = model.fit(train, epochs=10, validation_data=valid)

It works fine, but if I uncomment metrics line, I get this error:

ValueError: as_list() is not defined on an unknown TensorShape.

If I use objects instead of strings ( metrics=[Accuracy(), MeanIoU(2)] ), it gives this error:

TypeError: '>' not supported between instances of 'NoneType' and 'int'

What am I doing wrong and how can I fix this?

UPD: If I use accuracy for both outputs ( metrics=[[Accuracy()], [Accuracy()]] ), I train without any error, so I conclude there is something wrong with MeanIoU in my code.

Btw, there is prediction for batch(8) samples (two outputs: class + coordinates as 4 numbers):

(array([[0.7866989 ],
        [0.973974  ],
        [0.9148978 ],
        [0.28471756],
        [0.9899457 ],
        [0.99033797],
        [0.7237025 ],
        [0.81942046]], dtype=float32),
 array([[0.2515184 , 0.25495493, 0.3642715 , 0.09299589],
        [0.87964845, 0.3134839 , 0.54833114, 0.36701256],
        [0.0304133 , 0.45813853, 0.19692126, 0.244534  ],
        [0.22500503, 0.70299083, 0.00123629, 0.41123846],
        [0.37099576, 0.6092719 , 0.13407992, 0.40188596],
        [0.32103425, 0.6240243 , 0.02281341, 0.03058532],
        [0.28678325, 0.19885723, 0.50342166, 0.57963324],
        [0.41590106, 0.21439987, 0.94105315, 0.3379435 ]], dtype=float32))

I thought may be format for MeanIoU is wrong, but arrays of 4 numbers seems valid for MeanIoU, doesn't it?

Answer 1

As I answered here , correct metrics are: BinaryAccuracy and custom MeanIoU ( default MeanIoU is not applicable to bboxes regression as I understood). Working code snippet is in the first link.