number 0 start from 204, from 0-204 is number 9, idk if its only me, but the data shows me the incorect label, for ex for y[0] it shows me 0 instead of 9
There seems to be some wrong labels in original Y data. It might be lowering the accuracy, isn't it ? Please check my codes below. plt.figure(figsize=(12,12)) for i in range(64): plt.subplot(8,8, i+1) visualize(X[i+290], label=y[i+290])
Wow, you are absolutely right! Apparently, this is the fixed version of the dataset: github.com/darkin1/sign-language-digits-ml/blob/master/dataset_fixed.zip Thanks so much for pointing this out! :)
Love the series! Really insightful vid : )
Thanks, Edward!
number 0 start from 204, from 0-204 is number 9, idk if its only me, but the data shows me the incorect label, for ex for y[0] it shows me 0 instead of 9
how to reshape the 64*64 into 28*28
suggest you add cmap='gray' in imshow otherwise might confuse folks
Thanks for the tip!
There seems to be some wrong labels in original Y data. It might be lowering the accuracy, isn't it ? Please check my codes below.
plt.figure(figsize=(12,12))
for i in range(64):
plt.subplot(8,8, i+1)
visualize(X[i+290], label=y[i+290])
Wow, you are absolutely right!
Apparently, this is the fixed version of the dataset:
github.com/darkin1/sign-language-digits-ml/blob/master/dataset_fixed.zip
Thanks so much for pointing this out! :)
@@gcdatkin I'm very thank you for your excellent video lectures !!!
No problem! I'm glad you enjoy them :)