Lack of inspiration. Helping scouts recruit the next big name.
What it does
Processes through .mp3 file and counts the number of occurrences of sports-related terms to eliminate the hassle of reviewing sports tape.
How we built it
We built the project using Typescript/Ionic/Cordova for the Android/IOS application, Python and NodeJS for the backend, and use Google Speech for Speech to text feature. To understand how our system works, please take a look at our system our architecture below:
Android Apps -> Full-audio translation system service (Python Backend) -> Google Speech API | V Streaming system (NodeJS Backend) | V Google Speech Stream API
Basically, we are using microservice concept to split the backend system based on the feature. Both services are deployed to Google Cloud. For our demo, we only use a local server for debugging purpose. To communicate to python server, we use standard HTTPS request. But for the streaming system, we use webRTC (socket). Ionic is our choice for the front end because we want to be able to support both IOS and Android with minimal effort.
Challenges we ran into
Lack of expertise in apache. We also had difficulty reaching mentorship for our project and had to do a lot of debugging on our own.
Accomplishments that we're proud of
We are all proud that we were able to work as a team and complete something that is usable and worthwhile.
What we learned
From converting speech to text, to setting up a python server, we learned a lot of little traits that helped put the project together
What's next for Text me by the hoops
- Complete the live streaming service
- Combine people reaction in a game and generate some kind of exciting animation effect
- Premium feature or Ads for business purpose
- Statistic and Analytic of the live stream and past stream