Generating two-digit images out of mnist dataset and then using CNN, CNN-LSTM models to classify them
In this project, I went through several steps including generating two-digit sequences of images and also implementing several models to classifiy them
The required steps for this project are as follows:
1- generating a two-digit mnist dataset ( to classifiy images into 11 classes (mnist has 10 classes and blank class)
2- implementing a CNN model (custom-model-with-batchnormalization)
3- implementing a CNN-LSTM model
To get an overview of the models' architecture, I have included the model visualizations in the Images folder. Needless to say, I visualized my models using Netron which is a viewer for neural networks.
- Image Classification
- Deep Learning
- Convolutional Neural Networks
- Recurrent Neural Networks
MIT License
Copyright (c) [2020] [Shakib Yazdani]
- Twitter - @iamshakibyz
- LinkedIn - shakib-yazdani