Differently Able

Inspiration

The belief that technology has the power to improve lives.

What it does

Ease of access has always been the priority of the people as the world is progressing on technology. Our platform provides easy accessibility to the so-called "Disable" people to be differently able using their hand gestures.

How we built it

We pretty much followed the Software Development Life Cycle with the prior ideas of our features, possible classes and then the programming and testing. We implemented Computer vision and Convolutional Neural Network using jupyter and Spyder to write the code in python along with the major libraries and frameworks like OpenCV, WX, Keras, and Tkinter. Moreover, the ideas came from different team members and we chose to go with the better one among those.

Challenges We ran into

Getting Training Data for the Deep Learning Model given that we had only one day to gather enough data, creating image data ourselves was very tedious and we couldn't find a much-related source for the data either. In addition, given the data size and the availability of the resources for the computation, the model could not be trained properly because of the time consumption that we had to reduce the data size by significant numbers that disturbs the symmetricity of the sample data.

Accomplishments that we are proud of

Got to develop something that we truly believe can direct us towards building to change lives and help the business world at the same time.

What we learned

Working with a team where everyone is focused to give their best makes developing more fun than playing FIFA.

What's next for Differently Able

The best part about Differently Able is that it is never done. There are always more areas to explore, from adding new gestures with complementing features to even just adding cleaner additional features and training data.