Speechtter

Inspiration

People across the world communicate using different languages and this should not be a barrier in accessing fun tweets across the world. We developed this application also keeping in mind people with disabilities of vision.

What it does

Our app allows for voice searching of tweets based on hashtags, and retrieves and reads out the most trending tweets in the language of the user. The users can search tweets in their spoken language as well.

How we built it

We decided to develop using Python due to the abundance of development packages. The speech_recognition package was used to get voice input, process it and convert it to text, which was then processed using the tweepy API to search for the relevant tweets. The fetched tweets were translated to the preferred language using the googletrans API. The translated tweet is converted to speech using the API provided by the gTTs and playSound packages.

Challenges we ran into

Collection of appropriate tweets using Twitter search API.
Preprocessing of the obtained tweets to a readable format
googletrans API did not always return results.

Accomplishments that we're proud of

Drawing inspiration from the workshop Intro to twitter API, we conceived this application and were able to put our ideas into an functioning application. Although building the application was time bound, through structured piece-mealed structured tasks, we were successfully able to work the model as a team.

What we learned

We got to learn about twitter API and libraries and how they can be used to fetch tweets based on different parameters. We also learnt how to use voice recognition and language translation using different packages and how important it is to perform preprocessing of tweets to make the tweets readable.
The workshops organized were informative in guiding us through the process.