Inspiration
As high school students in the very competitive Silicon Valley, we get out fair share of stress. It was important that we find an outlet for the stress; thus Talk2Me was born.
What it does
It uses Google's Speech API to interpret what the user has said, and then it responds using the vast databases it has access to.
How we built it
We used python to come up with the Tone Analyzer and used it to integrate Google Speech Recognition API into our code.
Challenges we ran into
We used the Anaconda developing environment, specifically, Spyder. We had two different python versions running on the same computer; 2.7 and 3.6. This resulted in "UnknownModuleError" for multiple people. Also, we used the wxPython library, and the calling that we were using was only compatible with 2.7. So when we upgraded in the middle of the hackathon, none of our modules could be found. It took about 45 minutes to an hour to diagnose those problems and treat them accordingly.
Accomplishments that we're proud of
This was out first Computational Linguistics hackathon, and after a long 8 hours, we finally cracked the Tone Analyzer and had it working at peak efficiency.
What we learned
We learned that we need to be very observant about the environments that we use to develop and the little nuances within them.
What's next for Talk2Me
We want to expand it and make it more user-friendly, and give it more resemblance to Siri and Cortana, some top-of-the-line virtual assistants.
Sources:
- pythonspot.com/en/speech-recognition-using-google-speech-api/
- tutorialspoint.com/python/python_while_loop.htm
- wxpython.org/pages/downloads
- python.org/downloads
- developer.wolframalpha.com/portal/myapps/
Log in or sign up for Devpost to join the conversation.