Conceptualization

With Project Oxford, I saw a good opportunity for Just Say It's concept. With the speech capabilities, there was a lot that could be done. However, I wanted to keep stuff simple since I didn't know much. I thought, "How many times does a language barrier get in the way of communication? Sure, they could type it in onto a translation service somewhere online, but what if we could make that faster?" That thought lead to Just Say It.

What Does It Do?

Just Say It is intended to function as a translator. It takes a foreign language (as speech input), translates that to English (in text), and then the device speaks the text in English. However, this version only takes text input, and can only translate from Spanish to English.

How It Was Built

I utilized Project Oxford APIs in conjunction with Microsoft Azure's web app hosting. I wrote it using Python, HTML, and a little bit of JavaScript.

Challenges and Adversities

I'm not very comfortable with using Python or JavaScript in a "real-world" situation: my only experience with Python is inside of my college class. I'm an incredibly inexperienced programmer, I have very little knowledge besides the basic datatypes. Additionally, I was unable to implement a speech-to-text function, multiple languages, and then a text-to-speech function due to insufficient documentation (Project Oxford text to speech) and a lack of understanding of Python, JavaScript, and APIs.

Accomplishments

I was able to submit something at the very least. Despite my team breaking up into smaller groups and working on multiple projects, I was able to construct this by myself. I had lost about 12 hours but I still managed to pull this together.

Lessons Learned

I learned that I need to voice my opinion more; if my team didn't want to go with my project, I should've started working on it from the start. Additionally, I learned a lot about using APIs, using Google to teach myself how to code, and working with a team. Despite some hindrances, a dedicated team (or person) can still recover and create something.

HackUCI is done. Now what?

Hopefully, in the future I can finish this and then port it out to mobile apps. People often have a hard time due to language barriers. Just Say It should be able to break down those barriers.

I worked with Tim Reilly from Microsoft and a fellow HackUCI peer, Ghiyath Alazzah, to implement text-to-speech via Project Oxford, and he worked diligently to help me get this function. Their code is linked in the second Github repository, and it will be implemented in the future.

Share this project:

Updates