hfawaz/dl-4-tsc

Issue using class activation map

paulomaia20 opened this issue · 4 comments

paulomaia20 commented

Hi,

Thank you so much for sharing the code.

I'm having some issues with using class activation map on a dataset trained using FCN:

x_train.reshape(x_train.shape[0], x_train.shape[1], 1)

ValueError: cannot reshape array of size 1272000 into shape (424,750,1)

My x_train has the size (343, 750, 4) in which I have 343 samples with 750 points and 4 channels. What am I doing wrong?

Thank you in advance

paulomaia20 commented

I managed to change the "1" to "4" in that code line and also changed it to 4 in:

ts = ts.reshape(1,-1,1) to ts.reshape(1,-1,4).

Is the obtained class map representative of all channels? I cannot obtain a class map per channel due to the architecture of the network, am I correct?

hfawaz commented

Hello,

Thanks for the interest in the code.

You are right, I think the changes that you have made are correct.

And yes, you cannot have a class activation map for each channel with the current architecture.

If the problem is resolved, feel free to close the issue.

Thanks

paulomaia20 commented

Thank you!

sadransh commented

@paulomaia20 Did you solve this for MTS? Could you please share how?

@hfawaz Could you please let me know how I visualize CAM for MTS?