Convnet | Python - Numpy Convolutional Neural Network | Machine Learning library

Thank you @Bas! I tested your suggestion on a Linux machine: for a file with ~239 million lines it took less than 1 min. By adding >lines.txt I could save the results. Interestingly, my first readLines R script needed "only" 29 min, which was surprisingly fast compared with my first experience (so I might have had some problem with my Windows computer at work which was not related to R).

In layer 3

Solved using this API modification in Sequential.cs:

Error is very simple .Its saying instead of 1 channel you have given 3 channel images.

one change would be in this block

As of April 19. 2020 it is not possible with the .NET Wrapper as documented in this issue on the GitHub page for Keras.NET

Your training images are greyscale images. That is, they only have one channel (as opposed to the three RGB color channels in color images).
It seems like your Dataset (implicitly) "squeezes" this singleton dimension, and instead of having a batch of shape BxCxHxW = 16x1x64x64, you end up with a batch of shape 16x64x64.
Try:

np.stack and np.array provide exactly the same array, unless you pass a specific axis to the second one.

Let us look at a smaller example on a tiny list of 2d arrays

The minimum is not a single point but as a rule a larger area. argmin finds the first x and y (top left corner) of this area:

In case of multiple occurrences of the minimum values, the indices corresponding to the first occurrence are returned.

What you need is the center of this minimum region. You can find it using moments. Sometimes you have multiple minimum regions for instance in frame107.png. In this case we take the biggest one by finding the contour with the largest area.

We still have some jumping markers as sometimes you have a tiny area that is the minimum, e.g. in frame25.png. Therefore we use a minimum area threshold min_area, i.e. we don't use the absolute minimum region but the region with the smallest value from all regions greater or equal that threshold.

It seems the batch_size should be 20 not 32.

Since you have steps_per_epoch = 100, it will execute next() on train generator 100 times before going to next epoch.

Now, in train_generator the batch_size is 32, so it can generate 2000/32 number of batches, given that you have 2000 number of training samples. And that is approximate 62.

So on 63th time executing next() on train_generator will give nothing and it will tell Your input ran out of data;

Ideally,