Walkthrough on how to fetch, load, and pre-process image dataset for Deep Learning by Raviraj Dasari

This video is specifically made for Deep Learning Bootcamp learners. Hence, the scope might seem limited.
Notebook for this video can be found here: github.com/dph...
------
Connect with us-
Website: dphi.tech
LinkedIn: / dphi
Instagram: / dphi.tech
Facebook: / dphi.tech
Twitter: / dphi_tech

[@momoben364]

Guys you are amazing ! Please continue the great work ❤

[@talalzafar1111]

Thanks for this amazing video
All concepts regarding data pre-processing cleared!! :)

[@aiplanet]

Glad to know it was helpful, thank you very much for sharing your experience.

[@user-mw1bx3pu7l]

Really very amazing explainations ❤

[@tekniyapa748]

thanks alottt

[@pritamrajbhar9504]

nice explanation.

[@joshiji2430]

thanks for the nice video

[@nikependragon]

input_shape was never defined. This tutorial cannot be completed.

[@tejaswivegi2131]

What should be done when the model is not predicting properly?

[@saiprasanthpaladugula9552]

what we have to do if our dataset dont have class names like cat or dog only it have images of both of them without class names then how to do preprocessing for that type of dataset?

[@tejaswivegi2131]

Hello, did you find any way for this?

[@saiprasanthpaladugula9552]

@@tejaswivegi2131 No, Actually I thought that I will get reply from the dphi people but didn't get reply yet if you find the way please let me know.

[@tejaswivegi2131]

Sure

[@ankitchaturvedi9453]

Hello Raviraj, will the simple ANN accept image as input? I got the shape mismatch error while using the below code.
model = tf.keras.models.Sequential()
model.add(tf.keras.layers.Dense(1024, activation='relu'))
model.add(tf.keras.layers.Dropout(0.2))
model.add(tf.keras.layers.Dense(2,activation = 'softmax'))
input_shape = (32, 100, 100, 3)
model.build(input_shape)

[@studology67]

First flatten it, before adding any dense layers

[@gailwittichlive9593]

a bit confusing - using the word 'label' to refer to a file name (rather than a Target).

[@aiplanet]

thank you, Gail, for sharing your valuable feedback. We will see if we can incorporate this feedback and add some descriptions to it.