The inspiration behind Translatte was that we realized language barriers are something that still divide us today; although English has become the global standard language we felt as though the diversity in the world should not be lost. That's where the idea of translation sprung. Once we thought about translation we thought about how we could develop something more than just a basic translation app, with a strong desire to acquaint ourselves with the Google Cloud Platform we realized that the Vision API provided Optical Character Recognition capabilities. That's when we knew we'd incorporate OCR as a means to pull words in for users off of printed text, instead of having them type the words in.
As we hacked, our minds began to wander and we thought about how great this app could really be, not just translation but a tool for travelers embarking on adventures in new countries. A tool for students who don't want to read a paragraph of text from their textbook, and would instead rather have it read to them. That's where the speech came to play, with the use of Android's text to speech capabilities we could turn any block of text into a miniature audiobook for the user to listen to.
Translatte feeds off the powerful capabilities provided by the Google Cloud platform. With the Vision API text recognition was great, and with a translation API stocked with numerous languages we were able to convert from english to many different languages and vice versa.
The first step in implementing these well built APIs was to get familiar with them, we set apart tasks for each team member and dove into an hour long study session where we walked through Google tutorials and got a better understanding for the code libraries. Once we understood what was happening we were able to work towards building our application.
Numerous challenges had began to arise, the translation API did in fact cover many languages but Android's built in text to speech could only read so many aloud. We had to find the overlap and use that as our set of target languages.
The Vision API was a bit of a hurdle as well as it was so jam packed with capabilities, that reading through the code and fully grasping what was happening without getting lost was an issue; however, by being patient and building on what we learned from the tutorial we were able to move forward.
We finished our application as we had wanted on time! For a 24 hour hackathon just building this awesome project was a major accomplishment.
We made sure that we could finish with time to spare for quality assurance, knowing the true hackathon deadline we set one for ourselves a few hour ahead to allow for optimization, UI enhancements, and an overall sleek product.
Putting your minds together and working through stressful situations with patience and good team effort provides for the funnest project.