There are many sign language users in America, however, the amount of population that can understand sign language is scarce. Thus, this has inspired us to design a device that will translate sign language, so that sign language users and non-sign language users can communicate freely, without barriers.The setup of the system will consist of a video recording device, an arduino or raspberry pi, and a Gyrfalcon dongle for performing fast mathematical calculations. The sign language speaker can wave the signs at the camera. Then, using the embedded chip, the video frames will be parsed and, hopefully, preprocessed. Afterwards, the preprocessed frames/images of the sign language will be passed onto the Gyrfalcon dongle, which contains a neural net for deciphering the sign language image. Finally, the translation is displayed on an LED screen.
- Checkpoint 1 2/7/19
- Understand the SDK and perhaps initialize a demo program using the Dongle
- Checkpoint 2 2/14/19
- Look into a software CNN for sign language recognition
- Checkpoint 3 2/21/19
- Finish a python version for CNN
- Checkpoint 4 2/28/19
- Complete the set up physically
- Checkpoint 5 3/7/19
- Transfer CNN using the SDK to the setup
- Checkpoint 6 3/14/19
- Test the set up
- Checkpoint 7 3/21/19
- Debug
- Checkpoint 8 3/28/19
- Write up
- Checkpoint 9 4/4/19
- Prepare for presentation
- Checkpoint 10 4/11/19
- Checkpoint 11 4/18/19
Translate to other languages Possibly create a mobile app that displays the translations in wanted language