Lucy - Sign Language Tool

Inspiration

Inspired by IoT devices and our interest in helping others

Translates sign language to text/speech

Using Myo Hand Gesture Wristband,sdk, and our algorithm

Encoding the signal and finding an efficient algorithm

That we're proud of we're able to track signals to a degree of accuracy

What EMG is and how IoT can be applied to a wide variety of applications

completing the algorithm and making it more efficient, possibly developing an app in the future.

Learning how to use C++ on Visual Studio from three awesome computer and electical engineers, I was able to implement a pattern recognition algorithm (Dynamic Time Warping) that introduces a new front to adding custom made gestures to the Myo Armband. I helped the team develop a data collection platform which translated the EMG pulse signals from the armband to our program’s interface. We were then able to add a history of data for very common and simple hand gestures (numbers on one hand and single-hand ASL gestures like “okay” and “goodbye”. The part I am most proud of is the list of different mathematical analysis techniques that we went through in order to create a robust pattern recognition algorithm. Starting from sensor value correlation, variance study, and , we were able to build up to an impressive dynamic time series algorithm that identified a small range of gestures by matching the real-time sensor value array to a set of templates specific to each gesture.

Ali Abdelhamid
Simon Mekonen
Nicole Fella
Ernesto Ortega

Leave feedback in the comments!

Log in or sign up for Devpost to join the conversation.