Check failed: ShapeEquals(proto) shape mismatch (reshape not set) ..?

[Joshua Batty]

Hey pannous, I managed to train the model over night and now trying to run the test-speech-recognition.ipynb.

For some reason i'm getting the following

F0722 12:16:15.248998 2071900928 blob.cpp:455] Check failed: ShapeEquals(proto) shape mismatch (reshape not set)
*** Check failure stack trace: ***

Have you come across this before? Or know what might be causing that error in the script?

Also I had to change the following line of code from:
net = caffe.Net(model,weights)

to:
net = caffe.Net(model,weights,caffe.TEST)

Thanks.

[pannous]

Looks like caffe lost the ability to read monochrome images:
the is_color: false flag was removed

Investigating the options ...

[Joshua Batty]

Ok great, good to know it wasn't some error on my end. Looking forward to any potential solutions....

[Joshua Batty]

Hi Pannous, if we used RGB Spectrograms like in the image below would that be a possible work around?

[pannous]

Hi Joshua,

you probably solved this yourself by now. If not: Yes, the pipeline could easily be reconfigured to consume colored spectrographs. Do you have a data dump available or was that just an example image?

[sunshinehuo]

Hi JoshuaBatty! I have met the same error with yours,it shows like:
I0416 19:15:01.163847 31334 net.cpp:235] This network produces output prob
I0416 19:15:01.164002 31334 net.cpp:482] Collecting Learning Rate and Weight Decay.
I0416 19:15:01.164043 31334 net.cpp:247] Network initialization done.
I0416 19:15:01.164062 31334 net.cpp:248] Memory required for data: 545512320
F0416 19:15:01.288144 31334 blob.cpp:455] Check failed: ShapeEquals(proto) shape mismatch (reshape not set)
*** Check failure stack trace: ***

How did you solve it at last?Do you have any ideas?Thanks very much!