I have used mnist dataset and applied 5 layer Convolutional Neural Network. This implementation is similar to the LeNet5 implementation in Keras, except some of minor changes like instead of sigmoid and tanh activation I have used relu and softmax activations. Moreover the no of filters is also different to give a better accuracy. I have also used data augmentation on the train set. The dataset contains 60000 grayscale images and test set contains 10000 images. I have achieved accuracy of 99.6% on the validation data.
-
Notifications
You must be signed in to change notification settings - Fork 0
I have used mnist dataset and applied 5 layer Convolutional Neural Network. This implementation is similar to the LeNet5 implementation in Keras, except some of minor changes like instead of sigmoid and tanh activation I have used relu and softmax activations. Moreover the no of filters is also different to give a better accuracy.
silentkinght25/Recognising-Hand-Written-digits-LeNet5-Implementation-in-Keras-
Folders and files
| Name | Name | Last commit message | Last commit date | |
|---|---|---|---|---|
Repository files navigation
About
I have used mnist dataset and applied 5 layer Convolutional Neural Network. This implementation is similar to the LeNet5 implementation in Keras, except some of minor changes like instead of sigmoid and tanh activation I have used relu and softmax activations. Moreover the no of filters is also different to give a better accuracy.
Resources
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published