I realized that in the example notebook, when you try to open the tensor storage, the whole image slices (result) passed as the depth, and not just those where selected (src). This caused to append a large storage space to the one created by the inference process.
So, I just added the srcD variable, and used it to open the right size storage.
I also add some print log to the inference process that might help to see where we are in the process.
Hi, nice work!
I realized that in the example notebook, when you try to open the tensor storage, the whole image slices (
result) passed as the depth, and not just those where selected (src). This caused to append a large storage space to the one created by the inference process.So, I just added the
srcDvariable, and used it to open the right size storage.I also add some print log to the inference process that might help to see where we are in the process.