Character recognition is the conversion of printed or handwritten text (present in an image) into machine – encoded format, so it can stored efficiently, searched and edited more quickly or used as an input to text mining and other such applications.
For this project, I will be using Google Colab. For those who are unaware, it is a research project created to disseminate ML education and you can train your neural networks on their Tesla K80 GPUs. It sure speeds up the entire training process. Do check it out.
Make sure that you install the following libraries, if you are running it on yours system:
- matplotlib
- numpy
- sklearn
- keras
- seaborn
If you are interested in modifying this model and training it yourself, then please feel free to do so.
i. Download the IPython Notebook.
ii. Open google Colaboratory. / Run Jupyter notebook.
iii. Upload this notebook from your system.
All the details regarding usage of the dataset is given inside the iPython Notebook.