Our idea sprouted from the difficulties of public speaking. Every day students, teachers, politicians, and even regular people must speak in front of their peers and strangers alike. This brings a tremendous amount of pressure to perfect their public speaking skills. Our program aids with this process; "F Speaking*" allows users to practice their presentations using correction algorithms comparing what the user wants to say (which they post into the software) and what they actually say.
How it was built
By utilizing node.js as the back-end to integrate Google Cloud Speech to Text API, we are able to capture your voice and transcribe it into text. Through the use of a comparison algorithm and HTML, CSS, and p5.js for front-end design, our product delivers the elegance and simplicity yet also being extraordinarily powerful for everyday use.
The integration of Google Speech to Text API through node.js was one of the most difficult challenges we faced yet. The ran into countless failures until success to integrate the API with our design. Furthermore, the comparison algorithm was no easy feat either. The algorithm was modified until a functional prototype was developed and created.
Our group achieved many accomplishments that we are incredibly proud of. One of which is the huge amount of back-end programming that we accomplished plus the great amount of node.js we learned over this weekend.
Our next plans for "F Speaking" will be the integration of even more features such as correction of grammar/punctuation, proper speech enunciation, and even integration of machine learning AI to further enhance the capabilities of this program.