SLR - Sign Language Rocks

• 

  Hardware Design

• 

  Capturing letter "L"

• 

  Capturing letter "O"

• 

  Capturing letter "E"

• 

  Capturing letter "V"

Inspiration

Over 70 million people around the world use sign language as their native form of communication. 70 million voices who were unable to be fully recognized in today’s society. This disparity inspired our team to develop a program that will allow for real-time translation of sign language into a text display monitor. Allowing for more inclusivity among those who do not know sign language to be able to communicate with a new community by breaking down language barriers.

What it does

It translates sign language into text in real-time processing.

How we built it

We set the environment by installing different packages (Open cv, mediapipe, scikit-learn), and set a webcam. -Data preparation: We collected data for our ML model by capturing a few sign language letters through the webcam that takes the whole image frame to collect it into different categories to classify the letters. -Data processing: We use the media pipe computer vision inference to capture the hand gestures to localize the landmarks of your fingers. -Train/ Test model: We trained our model to detect the matches between the trained images and hand landmarks captured in real time.

Challenges we ran into

The challenges we ran into first began with our team struggling to come up with a topic to develop. Then we ran into the issue of developing a program to integrate our sign language detection code with the hardware due to our laptop lacking the ability to effectively process the magnitude of our code.

Accomplishments that we're proud of

The accomplishment that we are most proud of is that we were able to implement hardware in our project as well as Machine Learning with a focus on computer vision.

What we learned

At the beginning of our project, our team was inexperienced with developing machine learning coding. However, through our extensive research on machine learning, we were able to expand our knowledge in under 36 hrs to develop a fully working program.

Built With

• arduino
• machine-learning
• mediapipe
• pycharm
• python