Inspiration
The inspiration comes from the desire to lift the Amharic language to the digital age.
What it does
It digitizes scanned unicode Amharic text into digital characters.
How I built it
I will build it using python and tensorflow deep learning algorithm.
Challenges I ran into
• Limitation on python language. We have basic python proficiency. • Getting large samples of handwritten texts.
Accomplishments that I'm proud of
• Having the ability and ingenuity to use machine learning to solve fundamental problems. • Showing that digitizing Ethiopic language is an achievable task.
What I learned
• How to install tensorflow using pip. pip3 install --upgrade https://storage.googleapis.com/tensorflow/mac/cpu/tensorflow-1.0.0-py3-none-any.whl • That a tensor is a multidimensional matrix. • Although hundreds/thousands of written natural languages exist, there is distinct similarities among them.
What's next for Unicode (Amharic)Text Classification System
To establish free and paid web-based document converting service powered by Django 2.0. Scalable web app
Log in or sign up for Devpost to join the conversation.