ValueError: need more than 1 value to unpack

[wanghuok]

Hello, when i run 'Run detection on example' section in "ResNet50RetinaNet - COCO 2017.ipynb", i encounter this error, what's wrong whit it?

ValueError Traceback (most recent call last)
in ()
----> 1 image_batch, outputs = next(test_generator)
2 image = image_batch[0, ...].copy()
3 image -= min(image.flatten())
4 image /= max(image.flatten())
5 image *= 255

/home/wh/.local/lib/python2.7/site-packages/keras_retinanet/preprocessing/coco.pyc in next(self)
138 # lock indexing to prevent race conditions
139 with self.lock:
--> 140 selection, _, batch_size = next(self.index_generator)
141
142 assert(batch_size == 1), "Currently only batch_size=1 is allowed."

ValueError: need more than 1 value to unpack

[hgaiser]

What version of keras are you running ?

print(keras.__version__)

[wanghuok]

print keras.version
2.0.9
python 2.7.13

[hgaiser]

You can try this PR, or install keras 2.0.8. I suspect that will help with your issue.

[wanghuok]

Thanks, when i reinstall keras 2.0.8. that error was disappeared.
if i only want to run "Run detection on example" section, whether should i run "Train on data" section first?

[hgaiser]

If you load a model in your notebook then training isn't necessary.

[wanghuok]

Yes, i load a model "resnet50_coco_best_v1.0.0.h5".
but, when i run "Run detection on example" code, i encounter this error: "AttributeError: 'NoneType' object has no attribute 'shape'" at this line "image_batch, outputs = next(test_generator)". it seems that it didn't find any image. but my test_generator initialized successfully. please see below:

create a generator for testing data

test_generator = keras_retinanet.preprocessing.coco.CocoIterator(
'/home/wh/Downloads',
'val2017',
test_image_data_generator,
)
it outputs :
loading annotations into memory...
Done (t=0.54s)
creating index...
index created!

[hgaiser]

What is the structure of your COCO dataset? It should be:

/path/to/COCO/
/path/to/COCO/annotations/
/path/to/COCO/images/train2017/
/path/to/COCO/images/val2017/

[wanghuok]

Yeah. it works. thank you.