Hey, I am relatively new to Machine Learning and am trying

Question

Hey, I am relatively new to Machine Learning and am trying

to build a character recognition model to get a hang of it all. I am still stuck in trying to fit the model. What I did so far is, I have a directory named TrainImg which has directories named after the characters they hold. For eg: the first directory is named 0 because all the .png files in it are different handwritten zeros and so forth. I read from the directories and made a list of numpy arrays of the images. Also a list of labels corresponding to the directory. Then I converted both the lists to numpy arrays. The images are 900x1200 pixels. Any idea how to go about training them? I made a model with two dense layers the first one has a input shape of 900,1200,1. I don't know what the first arguments supposed to be, the book I was referencing had something like dense(512, input_shape(900,1200,1)). Is that how it's supposed to work? Any help would be great.

0

11.10.2018

1 ответов

15 просмотров

T-break · Accepted Answer

T-break

Which book is that ?

0

12.10.2018

Похожие чаты

Hey, I am relatively new to Machine Learning and am trying

1 ответов

Похожие вопросы