Week 2 (10/05/2020)

THIS WEEK

Tyler

• Laid out the initial design of the glove's hardware. As of now, each glove will have an MCU with six flex sensors (one for each finger, one for wrist bending) and an IMU (Inertial Measurement Unit) to collect data on finger and hand movements.

• Found Raspberry Pi 4 is the best MCU for the purposes of this project. It has 4 GB of RAM and a 1.5 GHz Processor. If more processing power or memory is required for machine learning, a 'Google Coral' AI accelerator may be used.

• Found flex sensors that would work well for detecting finger motion and wrist bending.

• Found a gyroscope and accelerometer in one board that would well for detecting the entire hand's movement.

Mark

• Reviewed gTTS and Pyttsx3 (Text-to-Speech). gTTs requires an internet connection to access the Google API to translate. Reviewed Google Translate as a possibility of translating different languages.

• Ordered Hands-on Machine Learning with Scikit-Learn, Kera & TensorFlow, and have began reviewing the book.

• Browsed online for American Sign Language (ASL) phrases to be used. The American Sign Language Phrase Book 3rd Edition by Lou Fant and Barbara Bernstein Fant online, and other online resources regarding ASL and illustrations on how to perform them were found.

NEXT WEEK

Tyler

• Further review hardware for the gloves (i.e. transmitter).

Mark

• Continue to review Machine Learning textbook and work on the example code provided within the textbook.