Aadesh Neupane

Thanks @Kautenja.

Thanks @sgnet91. I tried finding the code I used to train the model. As its been years, unfortunately I couldn't find it. As far as I remember, I used Lecun...

Thanks @DecentMakeover . I will write a short description about using the scripts. Yes, you do have to train the model in keras.

When you call create_model function, it will create the distance.npy and mean.npy. I am writing a readme file to make it easier to use.

Thanks @DecentMakeover . Finally, I have updated the repo with detail to how to implement with your own model of choice and data. Hope this update will be useful.

Thanks @AvrahamTs . This code was a port from original paper implementation. In the authors original implementation, they used the factor 1/200 and I am just using that factor. Hope...

@Amizhthan171 . I think implementing openMax with PyTorch should be possible. Currently the project uses Keras but I believe it is possible to convert each functions to PyTorch. You are...

@Amizhthan171 , I am not familiar with LayoutLM so I do not know how we can use this repo would be useful. But based on my understanding, since we are...

Thanks @zhyhy . Did you try training a different network on MNIST dataset and evaluate the performance? For me, the classification accuracy was around 97%.

Thanks @sebanob77 . I couldn't find the original code for MNIST_CNN.h5. Based on what I remember, I wrote a Lecun net to train the MINST data.